Software Blog (beta)

Decoding BEA's stock

Some years ago, when I carried a bag as a salesguy, I noticed that in the short term there wasn’t a strong correlation between my firm’s performance in a quarter and our stock price. To be sure, the stock usually did go up on the strength of a good quarter or down on weakness, but there were plenty of times when the opposite would happen. It was frustrating to have a good quarter and see the stock actually go down the day earnings were announced! 

Take BEA Systems last quarter. They actually posted revenue right at the top end of the guided range, reporting total revenue of $392M against guidance between $378M and $392M. The license revenue guidance between $159M and $176M implied a midpoint of $168M. BEA actually printed $169M, which was also in-line with analyst’s expectations.

So what happened? The stock closed at $13.20 and then dropped to 12.40 in the aftermarket when the numbers came out. That’s a 6.1% drop for meeting stated guidance. What’s worse is that the stock dropped another 3.6% in the next day’s trading, closing at $11.95. All told, BEA lost 9.5% of its value. Put another way, BEA lost about a half billion in market cap, dropping from $5.3B to $4.8B.  

So what REALLY happened? Obviously the market didn’t like something it saw or heard. And to understand that, one needs to figure out the key metrics that investors follow to determine if management’s stated growth strategy is working or not.

There are many data points provided in a firm’s financials, but in BEA’s case there are six or so that investors place a higher priority on than others. They are:

1. License Sales
2. Total Revenue
3. Operating Margins
4. EPS
5. Deferred Revenue
6. Backlog
7. Cash Flow from Operations
8. Guidance for the next quarter and full year

• License revenue is pretty obvious: it sheds light on whether or not customers are buying new product. Exceeding investor expectations is good news for the stock. The opposite is also true. You find expectations by looking at a bunch of sell-side models and calculating the average estimate for license revenue.
• For BEA, total revenue paints the rest of the picture which is primarily that of maintenance growth. The expectation for this number is easy to find – go to Yahoo Finance and click on “Analyst Estimates.” The name of the game is to meet or beat expectations.
• Operating margins – this is how investors get a handle on how BEA is managing expenses. The company runs in excess of 20% historically and investors want to see this number go up over time. The company has made some promises about expanding this number, but what’s important to notice is how the # is trending. Up is good. Down is bad.
• EPS. So many institutional investors look to this metric that, agree with them or not, it is hugely impactful to the stock price. Once you have the revenue and operating margin, you have the revenue/expense story… EPS takes that and factors in all the other stuff like taxes, interest income, and so forth. The key with EPS is to see if the company exceeded, or disappointed expectations (which can be found on Yahoo Finance) and if they did so was it a one-time event or was it something that will likely happen again in the future?
• Deferred revenue and Backlog. Investors pay attention to this metric because this gives some headlights onto what sort of maintenance revenue impact will be. It also gives some sense for how BEA is tucking away license revenue into backlog.

Backlog is a funny concept – it isn’t a hard and fast GAAP rule. Every company out there has their own interpretation of how to treat backlog, but as long as they are consistent they are perfectly legal. To BEA’s stock, Backlog represents something like a piggy bank where some license revenue goes that isn’t recognized as revenue or put in deferred. Whatever BEA’s conditions are for determining backlog, it ultimately represents the BEA’s ability to smooth out quarters. No backlog means no cushion. When deferred revenue goes up, investors assume backlog also goes up a bit. 

You’ll notice I didn’t say that “deferred revenue equals backlog” but instead implied the two are correlated.

• Cash Flow from Operations – this is obviously important but hard to predict because quarter over quarter so many things can impact cash flow that it’s probably more useful to look at      it on an exception basis as well as a historical trend basis. In other words, if this spikes or      plunges, that’s going to hit the stock price. And if it is consistently going up or down that will also drive a stock up or down. The stock price doesn’t track against quarterly gyrations – it tracks to the long term trend.
• Guidance. Usually overlooked by the retail investor, this is crucial to how the stock should perform in the future. You can find what Wall Street is expecting for total revenue and EPS pretty easily – check Yahoo Finance for “analyst expectations” and there’s your quick ‘n’ dirty gauge.
• Day’s sales outstanding (DSO’s) – if they spike, then the quarter was really back end loaded which is bad. Or they’re being forced to offer extended payment terms to close deals which is bad, or collections could be bad. The opposite is also true.

So what happened to BEA? In a nutshell, they printed great revenue numbers as well as very strong deferred numbers. But the guidance was disappointing in that most investors were expecting something around $140M and instead the company guided to something around $134. It doesn’t sound like much, but it was enough to tank the stock.

Ray Ozzie Has Been Assimilated

Unbelievable.  According to Computerworld, Ray Ozzie was quoted at Goldman Sach's tech conference as saying,

"Ever since the [dot-com] era, technologists have been trying to see how much of the Office experience we could take up into a browser and make usable in some form," Ozzie said. "Yes, there's Google Docs & Spreadsheets, there's ThinkFree, Zoho -- there are a variety of different instances of people taking the tools and kind of replicating them up into a Web environment... In the pure Web model, the trade-offs are fairly substantial. You have to be online to use them."

This runs counter to the November 1, 2005 internal email from Ozzie and Gates stating that the next disruptive wave in computing was the "Services Wave" and they better get it right or else.  However, Microsoft are now singing a different tune.  To clarify his earlier comments, Mr. Ozzie said,

"The way that I view the services opportunity related to productivity is really about more than just taking what's on the PC and putting it up on the Web.  I think there are high-level scenarios that if you consider you've got software on PCs, services in the cloud and devices -- mobile devices -- as the power that you can work with, and you try to envision the value of productivity and what you're trying to offer, you end up with a different result."

Ok, enough bluster on my part.  What this is really telling me is that the Microsoft Live strategy will integrate tightly with the desktop applications and SmartPhone technology.  Taking shots at Google by saying Docs and Spreadsheets are weak because you have to be connected is just plain silly.  Wireless connectivity is ubiquitous.  (Now - if I am in Outer Mongolia with my laptop, then I might want my Excel in order to do my DCF's, but that is a remote scenario!) 

Reading the tea leaves on this, I feel even better about Salesforce.com.  I have been a bit concerned about the entry of CRM Live as it would compete squarely with the SMB business that is the lifeblood of Salesforce.com.  Now it appears that Microsoft is trying to a) tie in their legacy desktop strength, and b) add in cool new features such as mobile devices and "services in the cloud" for their Live strategy. 

In other words, they are trying to one-up the industry standard by out-featuring them while maintaining a strong umbilical cord to the core Windows/Office franchise.  In other words, they're building the Zune of CRM suites.

Keep buying CRM shares.

Oracle Buys Hyperion

Ending months and months of speculation that Oracle was going to buy a BI player, today the Redwood Shores software giant acquired Hyperion for $3.3B in cash.  This represents a 21% premium to the closing share price yesterday (Wednesday). 

Business Objects had been frequently linked to Oracle in takeout rumors and was down 5.6% in the pre-market trading.  Cognos, the other player in the space was down 0.4%. 

Oracle says that the Hyperion acquisition will help them further bring the fight to SAP's door.  Many of Hyperion's 12,000 customers use SAP and now that customer information is going into Oracle. 

Oracle's own BI product, itself acquired via Siebel, was a strong contender in basic BI functionality but lacked some key features such as Financial Planning - something CFO's are keen to have and something that Hyperion did rather well.  In that sense this is a good fit.

Oracle also stays on the offensive, and gives the wags out there something to write about other than their lackluster previous quarter. 

This also sheds light on the relative importance of BI versus infrastructure.  I find it interesting that Oracle still has not made a move on BEA but instead acquired Hyperion.  Granted, BEA would cost more but still the facts are the facts.  Oracle felt its BI needs were senior to its infrastructure needs. 

Oracle is trading off 0.9% on the news in the premarket.  Update to follow.

India Wrestles With Success

Fascinating article.  The author travels to Nasscon and notices, among other things, that Valentine's Day has taken hold in India, and the days of obtaining a room at a 5 Star hotel for $100-a-night are gone. 

The former point demonstrates an aspect of the cultural struggle going on within the subcontinent.  Money quote:

When I travelled to the Nasscom event last year there was an enormous row over Valentine's Day. It has only recently become a popular concept in India and angry mobs had burned shops that dared to sell gifts aimed at lovers celebrating the holiday - evidence of the local outrage at Western customs usurping local tradition. This is just a minor example of the struggles arising as India quickly changes and adapts to global customs and behaviour.

This year I saw billboard posters everywhere for Valentine's Day. Restaurants were advertising candle-lit deals for couples, phone companies were encouraging text messages of devotion and shops everywhere had succumbed to selling kitsch  at twice the usual price.

The latter point suggests that India may be toying with the goose that lays the golden egg.

A further example of how India is changing is the availability and price of good hotels close to where people want to do business. The main hotel chains no longer see any need to charge prices that reflect the fact they are located in a low-cost environment - they just charge international rates. I could not even get into the conference hotel for the second year running because it was booked months in advance - at something like $300 per room. Only a few years back I was enjoying the 5-star experience in India for $100 per night. Those days are truly over for any business traveller to India now.

But the most interesting and potentially impactful story is the tax story.  When the IT services industry entered it's hypergrowth phase, the government extended tax breaks to those firms until 2009.  Now that year does not appear that far off, and the companies are lobbying furiously to extend those breaks.  I am no authority on Indian tax code, but at the minimum, any tax increase will be passed on to the customers.  A company with a lush tax history like India's would give me pause.  Could they be that short-sighted?

In a development that warms my capitalistic heart, Dubai is offering 30 year tax havens for IT services to relocate or startup there.  Indian people are wise not to think of their success as inherently Indian, any more than automobile success in the early 20th century was inherently American.  Other nations can and will compete.

Looks like India will continue to be a relatively more inexpensive IT services region, but those savings are not going to be as significant as in days past.  Anecdotally I've heard that the cost of offshoring to India just isn't as cheap as it used to be, and companies in search of low IT services costs are scouring the former Eastern Bloc countries with some success. 

The Long and Short of SOA

Good article here on the challenges and payoffs of implementing a SOA strategy. 

The article does a good job communicating how significant a change SOA is to a company.
"You need to “turn your organization on its side” and incorporate SOA into thinking enterprisewide, said Steve Wolf, Marriott’s senior enterprise architect. “What separates SOA from a minor technology change is the fact that this is a fundamental change in the way we do business,” he said during a session at The Open Group LLC’s Enterprise Architecture Practitioners Conference.

Implications:
A) This is a long sales cycle... and therefore expensive
B) Without significant internal evangelizing and championing, SOA projects will fail to meet expectations

In other words, this sounds like many other "revolutionary" three-letter-acronyms: ERP, CRM, SCM...  Lots of promise but not easy to attain.  So far SOA does not have the stain of overpromise/underdeliver that the other three do, but that could change if the tech media gets a hold of the "SOA is a failure" meme. 

The article also does a good job of illustrating the benefits of all the hard work to implement SOA:
Con-way Inc. has seen a substantial return on investment from an SOA approach it has been using since 1998...  The IT department can now change business processes as soon as an alteration is needed instead of having to rewrite code... In addition, the SOA supports the electronic transmission of data to customs officials in Canada. That has slashed the average time it takes for one of Con-way’s trucks to pass through customs gates to enter Canada from two to three hours to less than a minute... (emphasis mine)

Nick Carr Was Right!

The following is an excerpt from a note I wrote a while back that was intended to educate investors to the reasons behind SaaS and Open Source.  It was never published and I just stumbled over it recently.  I thought this passage was interesting for the historical parallels.  While it is a strong endorsement of the SaaS model, I am not saying that SaaS will wholly replace traditional sofware delivery - but the trend toward sharing infrastructure that adds zero competitive advantage is economically wise and technically feasible.  Enjoy!

History Repeats Itself

At the beginning of the Civil War, there were 20 different railroad track gauges in the United States ranging from 3 to 6 feet.  The continent had developed several different lines in different geographies as the nation began to expand westward. The wisdom of the time held that railways were only practical where canals were not, and there were no plans made for the standardization of railway gauges. This proprietary infrastructure did not add any value to the freight company that used it, but it was essential for the service.  When freight items had to move from one railway to the next, they had to be manually unloaded from one to the other, due to the incompatible railway gauges.  The inefficiencies of this system gained a desperate quality to them when the Civil War broke out.  Suddenly, the penalty for delivering freight went up dramatically - battles were won and lost, partly due to the infrastructure behind moving armies and supplies.  The battles get the headlines, but the logistics significantly determine who fights who, with what equipment, and where.

With this urgency came the plan to standardize on a single gauge- 4 feet 8.5 inches.  During the war, over 4,000 miles of track was laid to this standard which ended the economic reason for any trains to be constructed for any other gauge.  The common infrastructure was shared by many competing railway firms.  Sound familiar?

Another historical example is the advent of the "Union Station."  First introduced in Indianapolis, Indiana in 1853, this was a common station that consolidated railroad passenger traffic.  By doing so, passengers could easily move from one system to another.  The alternative was having each passenger rail company build a station that would connect only with their own other passenger lines.  The real estate alone made this notion an economic non-starter.  Again, sharing the common infrastructure was the key enabler.  The parallels are obvious: the inflexibility, cost, and inefficiency of proprietary standards versus the ease of integration using agreed-upon standards and shared infrastructure.

Another way that history is repeating itself is similar to the path electricity took.  Thomas Edison's direct current (DC) standard was the format for transmitting electricity in the United   States.  This standard was sufficient for the loads at the time, but lacked the capability to be transmitted distances of over one mile without severe voltage dropoff.  A superior solution for transmitting electricity over distance came from Nikola Tesla, who called it alternating current (AC).  As Edison was loath to give up the royalties accruing to his patented DC system, he began an aggressive campaign to discredit Tesla's AC standard.  Edison went so far as to electrocute to death stray cats and dogs- even an elephant once- to demonstrate the hazards of the AC standard.  Despite this gruesome tactic, AC became the adopted standard for long-distance distribution of electricity.

This example illustrates a time when the standard for energy distribution made headlines, despite the fact that today we could not care less about electricity.  People only care about the services performed using electricity, be they appliances, devices, or whatever.  Electricity became a commodity that is simply thrown on or off as needed.

IT is the next utility

IT started out as something that every company built to support the business.  Today, this model is under considerable strain given that 85% of the IT budget goes to maintenance of the existing system.  Labor is wasted maintaining the current system, not creating innovative applications at the touchpoints with customers, employees, and/or partners.  The center of gravity in the IT shop is maintaining the status quo.  In other words, IT shops are spending the vast majority of time optimizing processes that do nothing to differentiate their business from the competition.

Hardware utilization hovers around 30% as IT purchases need to be prepared for the contingency of seasonal (or random) spikes in traffic.  The rest of the year, the hardware remains idle while costing the company monthly maintenance.  This over-provisioning is wasteful and inefficient.

With significant overlap of basic IT functionality shared by nearly all organizations, it is becoming more sensible to evaluate a shared infrastructure system.  On Demand software is exactly that model- a shared back-end system that allows companies to dial up their services as-needed.  This is the only way to achieve economies of scale in IT.

The notion of a Service Oriented Architecture (SOA) is not new.  Possibly the first book on it was by David Parnas, entitled "On the Criterion for Decomposing Systems into Modules," published in 1972.  The concept of SOA is to develop a technology backbone that allows services to be utilized in a "plug and play" fashion.  The entire system is designed for flexibility and lets all software assets get repurposed into new applications quickly.  A well designed SOA enables a firm to build out automated business processes rapidly, effectively reducing the time and cost of new initiatives.  The reason SOA is happening now is because of the popularity of the underpinning technologies.

The popularity of such standards as XML, SOAP, WSDL, and BPEL (to name but a few) is significantly impacted by the fact that these standards are not owned by any one vendor.  Code designed to these standards flourished and created a virtuous cycle where every addition benefited the entire group writing to these standards.  Thirty four years after David Parmas' book, SOA is hitting the mainstream because of the maturity of these standards.

Standards based communication and interoperability are crucial to SOA, but without an abundance of bandwidth to the home, SOA is restricted to individual firms, and utility computing is not feasible.  But with the surge in bandwidth comes the ability to render services to the market through a browser. Suddenly the software doesn't matter so much as the service does.

The commodification of much of the IT computing stack makes it ripe for outsourcing.  Without effective standards to enable simpler interoperability and communication, the utility model is not practical. But combining mature open standards with the availability of broadband makes it not only practical but compelling. 

The World Has Changed

The ubiquity of these standards and broadband has also greatly loosened another constraint to IT shops, namely the scarcity of qualified development resources.  In a post-internet world, it is possible to tap into software talent in India for development support. The significance of this reality has not been fully realized in the market.  If firms can reduce their operating expense by accepting a utility model and receiving their software as services (a la Salesfoce.com), more money is available to invest in applications and automating business processes.  In short, the doors to innovation would swing wide open but for the lack of developer talent in the US Market.  Bringing online qualified and inexpensive offshore developer talent will creates an opportunity for firms to create more value-added applications which should benefit their partners, customers, and employees.

This excess supply of applications will necessarily result in a slow decrease in pricing, assuming that the applications are written to standards.  This is a safe assumption since writing a proprietary application will be at a disadvantage in the market due to the headaches associated with integrating proprietary applications with other applications.

Red Hat Lives To Fight Another Day

No doubt due to the massive pressure put on them by Oracle and Microsoft, Wall Street fund managers had piled up 20.4 million shares short going into today's earnings announcement.  These same managers were forced to cover their bearish bets when Red Hat posted subsription revenue of $88.9M, up 48% annually, a positive surprise.  EPS came in at $0.14 compared to the Street's expected $0.12, and cash flow boomed at $59.7 million - well above the expected $45 million. 

Terry Tillman from Suntrust summed up guidance, "The company gave guidance for 4Q07, with total revenue in the range of $112 million to $113 million. GAAP operating cash flow for the full fiscal year is unchanged, suggesting a midpoint range of $41 million - $42 million, representing a significant downtick sequentially. Adjusted EPS is expected to be between $0.14 - $0.15."

CEO Matt Szulik was reserved in the face of some welcome good news for his company.  He attributed the success in the quarter due to the fact that customers don't find value in saving $5 per OS simply because some new player entered the market. 

To my mind, this is reminiscent of the outstanding/outrageous run of heroic quarters PeopleSoft put up in the face of Oracle's takeover offers.  In hindsight, the threat from Oracle focused the management at PeopleSoft to such a degree that sales execution was well above the norm.  It didn't hurt that PeopleSoft was also offering massive concessions to customers in order to get deals done.  It should be noted that PeopleSoft eventually drained the swamp dry, and finally imploded just prior to the board reversing their stance and accepting the Oracle bid. 

Jason Maynard from CSFB poked around this theme during the Q&A on the call, wondering aloud if perhaps there had been some special discounting going on.  Katherine Eggbert asked if special spiffs had been implemented to stimulate the bookings.  In both cases, management denied doing anything special and attributed the success to customer loyalty and a healthy demand environment.

I am not buying that at all.  There is no such thing as bad publicity, and the Oracle move certainly put the spotlight on Red Hat for a while.  That kind of brand awareness usually costs a pretty penny and Red Hat picked it up for free.  I also think that execution probably improved as the folks at Red Hat realize they are fighting for their lives.  Ultimately, I do not believe that discounts were not impacted by this.  Even if Oracle was a disaster in executing their support offering, I think savvy purchasing agents would leverage the heck out of the announcement to obtain more favorable pricing from Red Hat. 

In summary, we must acknowledge the strong numbers the company put together.  But even so, they will have to string together several more to make me a believer. 

The $1 or $10 billion company

Salesforce.com has earned a bit of press today, announcing that their forecasted FY2008 revenues would be higher than they previously estimated.  Last month, the On Demand CRM vendor forecast FY2008 revenues to be between $700 and $710 million, and today they have raised that to a range between $710 and $720 million. 

The subhead to this story is the release of the AppStore, which is alternately called the "iTunes for Salesforce.com" and the "Amazon.com superstore for Salesforce.com" in internal memos.  Basically this is a way for Salesforce to provision customers with the myriad applications within the AppExchange on a single Salesforce.com bill. 

What I like about this:
1) By creating a billing and collection service for AppExchange vendors, Salesforce.com is relieving ISV's of the burden of these onerous tasks.  In so doing, Salesforce.com eliminates a distraction from what they want their ISV's to do - namely innovate on their platform. 
2) A single bill is something customers want.  Think about all your cable, mobile, and long distance bills.  It's nice to have them all in once place.
3) This signals a strong focus on monetizing the AppExchange.  For Salesforce to go completely vertical in growth, they must leverage the Internet as a sales and distribution arm like no company has done in history.  AppExchange promises this, sporting over 400 applications today that are essentially off Income Statement R&D and S&M, while driving the overall value of the customers investment in the CRM application.  A converged billing solution shows foresight and attention from management in addressing this much larger opportunity.

What I'm not sure about:
1) While a converged billing solution is fantastic for everyone, it does not necessarily address another problem: micro-billing.  This is the situation unique to composite applications where an end user might be using web services from several different vendors.  Currently, each vendor bills for maximum usage and chokes off the ability of the end user from using more and more AppExchange apps.  What has to happen is a trusted and reliable way to send even micro-pennies to vendors based on the usage of their product, rather than force the whole cost burden upon a user.  This announcement suggests they are looking at this, but it lacks any detail around the solution.

What I don't like about this:
1) The cost.  The one way to screw this up to a fair-thee-well is to price the billing solution such that it isn't cost-effective for an ISV to use the system.  Right now, Salesforce could command up to 45% of the revenue from an application sold via the AppStore.  Why not encourage ISV's to populate the AppExchange by subsidizing the AppStore cost?  The network effect applies to the AppExchange, so why not do everything possible to enhance that value?  What you give away in revenue today is more than made up by the volume of tomorrow.

Risks:
Can Salesforce balance the demands of running a CRM company with that of blazing the trail of the eBay of On Demand applications?  The former strategy will get them to $1 billion in revenue, the latter $10 billion.  Why?  The direct salesforce can only scale so far so fast, and the market is big but not as big as the market described by the various applications in the AppExchange.  Furthermore, the ability of the AppExchange to address that market is similarly larger via the Web 2.0 buying methodology of "try it for free" and then feel the upward pull into paid-for and supported versions. 

The AppStore billing solution is set for December 2007, however the AppStore will launch in February of 2007.  Revenue estimates are not there yet, but need to be considered.

Conclusion:
Salesforce.com continues to be the most interesting enterprise software company around.  It will be instructive to see how Marc Benioff steers the course between being a new take on an old business (CRM) and creating something the world has never seen before (the AppExchange).  Based on what I've seen and heard thus far, I think they have a really good shot at achieving both. 

BPM, BPEL, SOA... who's doing what now?

Nobody is immune to it.  Eventually you're in a conversation somewhere and the acronyms are flying and the jargon is charging.  You are following along, nodding perhaps, and then in a flash... you've lost the plot.  The person on the other end of the phone is going on and on about a solution that was really cool and it used BPM, SOA, and BPEL.

Wait, you think to yourself, BPM is "business process management" and that's when they design business processes... like using Visio... is it Visio?  Note to self, check on Visio... anyway, BPM is hot, right?  Isn't everyone touting it like mad?  Who exactly is doing this design?  I've never met a "business analyst," have I?  How many business analysts are out there and could they possibly justify all the product hype? 

Meanwhile, the conversation is still rolling on.  This time it's about SOA.  Whew, you think to yourself, this is familiar ground.  SOA.  SOA isn't ESB, but ESB can be a part of SOA, right?  Or do you even need an ESB?  Haven't we had SOA around for a while?  I thought web services were the visible manifestations of SOA, or are they the other side of the SOA coin? 

Anyway, BPEL.  Isn't that the language for designing business processes?  But BPM does that.  Maybe BPM uses BPEL.  Or maybe they compete.  Do I need BPEL?  Does the hypothetical business analyst need it? 

Of course, your mind isn't this scattered (or is it!).  Nevertheless, this article clarifies all. 

BEA Q3 Earnings Report

Income Statement Summary

License revenue came in at 136.4 million, which was below consensus estimates of $140M but within management’s guidance of $130 to $146 million. Q3 license was up 30 basis points quarter over quarter and up 12.4% year over year. Total revenue was $347.6 million, up 2.4% quarterly and up 19.3% from a year ago. This met analyst expectations of $348.7 due to stronger than expected contribution from the services line.

Due to the ongoing investigation of potential options backdating, BEA was not able to release full quarterly financial numbers. Beyond the license and total revenue numbers, BEA did state that the “interest and other income” line was $9.5 million. During the call, management said BEA continues to improve the efficiency of their operations, implying improvement over last quarter’s 20.1%. Management also gave the pro forma tax rate of 33.5% and GAAP tax rate of 36%

If we assume a 20.6% operating margin, we arrive at operating income estimate of $71.6 million. To this we add the given “other income” line of $9.5 million which results in a pro forma estimate of pre-tax income of $81.2 million. Using the given pro forma tax rate of 33.5% we can determine the appropriate taxes of $27.2 million. Deducting this number from the $81.2 million, we find a pro forma estimated net income of $54 million. Divide this by the shares outstanding of 414.7 million and the pro forma estimated EPS comes to $0.13.

This estimated EPS is a penny below Wall Street’s estimate of $0.14 due to a higher than expected tax rate.

Balance Sheet Summary

BEA holds $1.406 billion in cash and short term investments, up from last quarter’s $1.368 billion. DSO’s came in at 71 days, up from last quarter’s 68.5 days.

Deferred revenues were down over $18 million, from last quarter’s $356 million to $378 million. This was attributed to the natural trend to see negative growth in deferred revenues from quarters one through three, and an increase in quarter four. The change in deferred was due to the timing of support contract renewals and the license component was negligible.

Cash Flow Summary

Cash flow from operations was up slightly from $50 million to $52 million quarter over quarter.

Guidance

Management guided investors to expect total revenues in the fourth quarter to be between $378 million and $392 million. License revenue was indicated to be 42%-45% of total revenue, implying a range of $159 million to $176 million. CFO Mark Dentinger pointed out that additional expenses will be incurred for the ongoing options investigation, somewhere around $2 million in the quarter. Furthermore, management indicated that there would be a one time expense of around $9 million paid out for employee bonuses.

The non GAAP tax rate was indicated to be 29.5%.

Stock Option investigation update

BEA initiated, under their own discretion, an internal investigation to ensure historical grants of stock options were not inappropriately backdated. No timeline for completing this action was given, although CEO Alfred Chuang indicated he hoped to have a breakthrough by the end of December.

Commentary

This was a disappointing quarter for the company, especially given the buzz around SOA and AquaLogic. While a $4 million license line miss isn’t egregious, the stock was down around 7% in the aftermarket, indicating investors had bid the stock up in anticipation of better news today.

What was buried in the call was a strange situation where a seven-digit deal was flipped from a license deal to a support deal in the 11^th hour by BEA’s auditors. Evidently a long time customer had a significant deal on the table with BEA and the auditors initially blessed the revenue as license revenue. After the deal was signed, the auditors reversed their opinion and categorized the revenue as support revenue which moved to the deferred line and not the license line.

Europe was also singled out as having a softer than expected quarter. The EMEA sales team is now headed by the former country head for Germany, a person who was responsible for turning that country around.

The tax rate was also higher than expected. Last quarter’s guidance indicated a 29% non GAAP tax rate, and this quarter produced a 33.5% rate.

The combination of the higher tax cost, the deal that flipped from license to support and the weakness in the license line produced an estimated EPS miss of one penny.

AquaLogic gained most of the good news spotlight, coming in at about 20% of license revenue. However, this is exactly the same way AquaLogic was characterized last quarter which implies AquaLogic grew the same rate as license overall – about 0.3%! Given the hype surrounding AquaLogic and SOA, we expected a higher percent contribution to license revenue.

This disappointing quarter raises some questions about BEA, who has been successful in portraying themselves as the pure-play leader in the SOA space. Does this signal execution issues at BEA or does it suggest a pause in SOA adoption?

Chasing deals

Standard article in BusinessWeek about a Web 2.0 bubble here, but I thought this statistic was worth mentioning:

"After a sharp decline in investments following the tech bust of 2000 and 2001, venture firms are beginning to pour money into a new crop of Internet companies, in businesses such as social networking and online video. Together, they're called Web 2.0 companies. Venture firms have put a total of $455.5 million into 79 of them during the first nine months of the year, according to a study released on Nov. 7 by market researcher VentureOne, which is part of Dow Jones & Co. (DJ). That's more than twice the amount of money that was invested in such companies during the same period in 2005."

Emphasis mine. 

Red Hat Under Siege!

In a move that I still can’t totally figure out – more on that later – Microsoft and Novell announced a partnership that resolves a potential dispute between the two vendors over intellectual property in Novell’s linux distribution. The announcement also touched on the two firms working together to enhance technical interoperability between the OS’es, and it mentioned that Novell would be paying Microsoft royalties for this “patent peace.”

The winners in this situation are the customers who have both SuSE linux and NT servers in their IT system. It eliminates the threat of a legal war erupting over who owns what IP, a fight that would both retard Linux expansion and do no favors to Microsoft’s image as a bully.

Novell is also a big winner here (stock up 14.65% at press time). They have managed to do about nothing relevant or interesting since Chris Stone went on the open source buying spree in 2003. New CEO Ron Hovsepian didn’t take long to shake things up at Novell and was credited during the Q&A as initiating the dialogues with Redmond. Suddenly Novell goes from circling the drain to the owner of the Linux distribution that has made peace with Microsoft. And that’s a huge differentiator.

So why did Microsoft do this? Because of the customers? Bah. Consider they could have done this a year ago but didn’t. The relevant landscape hasn’t changed much in that time. If you buy that, you have to explain the sudden change of heart.  And just how many customers are tearing their hair out because they can’t get their SuSE server to interoperate with their NT server? 

So why did Microsoft do this? Because of the royalty? Pshaw. A drop in the ocean to Microsoft.

So why did Microsoft do this?  A response to Oracle?  The irresistable urge to kick Red Hat while they're down?  Amusing to consider yes, but is that enough of a reason to legitimize the principle threat to the Microsoft cash cow(s)?

So why did Microsoft do this? Because they suddenly look in-step with reality for a change? That, combined with the previous two reasons probably gets us halfway there. But again, Microsoft had no visible compelling reason to do this TODAY. Deals just don’t get done unless they have to. And I am too dim to see why Microsoft HAD to do this.

There just doesn’t seem to be enough benefit to Microsoft to justify this move. They don’t need the money. Their customers are beholden to them already. Despite Linux’s runaway success and threat to their core business, Microsoft could chug away for years without a noticeable dip in their earnings.

The big loser, of course, is Red Hat. The Raleigh, NC outfit has taken two huge shots in the past few weeks from two of the heaviest hitters around. First Oracle threatens 95% of all their revenue by undercutting Red Hat’s support by half AND indemnifying users at the same time. Then Microsoft lends legitimacy and competitive differentiator to their biggest competitor. I can’t recall two such shots being absorbed by a single entity in such a short amount of time.

You know Oracle and Microsoft approached Red Hat. I can’t help but wonder what on earth the head of Business Development was thinking in failing to close these deals. Red Hat missed opportunities to partner with Oracle and/or Microsoft, eliminate IP issues, and put a stake through the heart of Novell. Of course we don’t know the offers extended to Red Hat, but as I mentioned in another post, they really didn’t have much choice in the first place. Red Hat is left with a Linux distribution that is cheaper to be supported by Oracle and isn’t indemnified by either Microsoft or Oracle.

Ultimately, what happened is that Microsoft has legitimized Linux, Novell is the principle beneficiary, Red Hat looks pretty bad, and Microsoft looks aggressive again.

Red Hat Defiant!

Matthew Szulik, CEO of Red Hat, essentially tells me to stuff it.  He is not, repeat not cutting prices to deal with Oracle's move. 

I do agree there will be loyalty to Red Hat, but at 2x the cost of a world-class support organization I wonder how long that will last.  I would suspect most people in charge of this decision will be asked to at least investigate what Oracle can do. 

Oracle Open World

Oracle Open World, 2006.  The acquisitions Oracle has made since 2004 took on a physical presence as 48,000 people attended the company's signature customer event of the year.  Oracle is a huge company, and this Open Word showed that all those acquired properties seemed to be coming together into something bigger.  But the big story of the week was Oracle's shot at Red Hat. 

The Big News
The Red Hat v. Oracle news is already old, with excellent analyses posted on Rough Type, Infoworld, and here.  Red Hat is furiously trying to spin it (check out their home page). 

What is interesting to me is why this even happened?  It is safe to assume Oracle was exploring doing this via a partnership with Red Hat.  So why didn't Red Hat do it?  Clearly someone in Raleigh (Red Hat's GHQ) didn't appreciate the threat Oracle posed to their revenue model.  Now substantially all of their revenue is at risk.  Red Hat did not have the cards to play, and needed to take whatever deal Oracle offered.  Their options now might include buying an open source DB like mySQL, but with their stock currency seriously devalued that gets harder.  Furthermore, mySQL relies on InnoDB for their storage engine - and we all know that Oracle bought InnoDB a while ago.  Right now, Red Hat is going to have to cut their prices to retain their clients, further weakening their position. 

About the only thing Red Hat can hope for - and this is largely out of their control - is that Oracle flubs it.  Oracle has a world class support organization but an OS is a different animal than a DB, and there will be a learning curve associated with scaling the support.

The big disconnect at Oracle is their view that this is not a new distribution of Linux.  Obviously they recognize the drawbacks in asking CIO's and ISV's to certify yet another linux distribution.  They strenuously pointed out this is EXACTLY the same as Red Hat's Linux.  Except without the trademarked Red Hat graphics.  And with some additional patches.  And then recompiled. 

Funny moment of the analyst meeting: The product manager for Oracle's Linux Support team was being very animated as he emphasized how this wasn't a new distribution.  The analysts weren't buying it at face value and someone whispered next to me, "It doesn't matter what you think... what do CIO's think?"  Yes, this passes for humor at these events...

It's not an Oracle meeting without a conspiracy theory!

 

The conspiracy theorists are now wondering if this is all part of a grand scheme from Oracle.  A plot so cunning and prescient that almost makes sense.  Let's see, Oracle destroyed $1B of Red Hat's $3.7B market cap the day after the announcement.  Red Hat has about $1B of current assets (cash, AR, ST securities, prepaid expenses) as well as $570M of debt on the balance sheet which gives you an enterprise value of about $3B.  The Oracle has over $8.3B in cash and short term securities alone.   

Assimilation Doing Well

I spoke with Oracle employees who were formerly at PeopleSoft, Siebel, and even JD Edwards.  OF course I chatted with Fusion reps, Apps reps, and Tech reps as well.  They all had been, well, assimilated (for lack of a better term).  I mean this is a good way - Oracle has matured into a customer focused, highly professional selling organization.  I was impressed at how the individual reps took advantage of the event.  Reps brought their customers in large numbers, held numerous appreciation events and leveraged the proximity of so many senior Oracle managers.  In other words, they did what they were supposed to do. 

Now this is much harder than it sounds.  You have to be on top of this 3-4 months out to simply get on your customer's calendar.  In most software companies I worked for, priorities one through nine were obtaining new license revenue.  Everything else, from doing expenses to helping the marketing guys with their data to getting customers to these events all tied for the 10th priority.  So when the event comes, it is swamped with people taking a boondoggle.  Everyone parties like mad, and not much real work gets accomplished.  The Oracle folks I met simply took everything more seriously and professionally.  Someone is managing this team well.  But I digress.

The point I want to make is that the conglomerate seems to be working out well.  The reps are more harmonious than I expected.  The customers were less skeptical about Oracle's support for their legacy systems.  The atmosphere was positive and Larry's strategy appears to be playing out as he intended.

The Road Ahead

Oracle has taken the haymakers SAP landed in the chaos surrounding the PeopleSoft takeover.  Oracle has been acquiring customers and maintenance streams, and the critical piece was to retain them.  This was in considerable doubt given Oracle's predatory reputation in the pre-Charles Phillips years.  But the partners and customers I spoke to all commented on how focused Charles Phillips was in talking to them, no matter how small or insignificant.  By extending the maintenance terms for all the acquired software products, Oracle eliminated any compelling event for customers to go to Microsoft or SAP.  By adding new features and functions to products that were starved by their original owners (I'm looking at you, JD Edwards) they won a critical mass of goodwill.  In short, Oracle did retain their customers and thus the ROI on these purchases looks realistic today.

With strong management and execution - two themes I believe I saw all week - Oracle is set to farm their huge installed base and keep those margins going north. 

The Instincts of a Monopolist Die Hard

I have to hand it to Microsoft.  Despite years of legal issues and the inevitable maturation brought on by the passing of time, they still bust out their basic (and illegal/unethical) playbook. 

This is the classic move where Microsoft allies themselves with a vendor who provides functionality they need.  They then reverse engineer the functionality and release a competing product that is inferior.  The future releases of the product access undocumented API's in the OS until magically Microsoft's product loads faster/performs better/whatever than the previous "partner" who gets aced out.

It's happening in Vista regarding security.  McAfee and Symantec are crying foul and howling.  The EU is interested.  And Microsoft is calling them - their critical partners - liars. 

My favorite anecdote to this story is the "glitch" which prevented the Symantec team from their meeting with Microsoft - is the Microsoft employee who set the meeting up a complete moron?  Is LiveMeeting that hard to master?  Should we all buy WebEx?

Hedge Fund Activism

I attended a seminar yesterday afternoon that took a long look at the recent spike in proxy fights.  According to Paul Schulman, three years ago there were 15 proxy fights.  Last year there were 26, and this year there are already over 50.  Mr. Freeman further went on to say that 60% of proxy fights wrangle some sort of concession from management, be it a board seat, a dividend, or a buy back. 

The spike in this activity comes almost exclusively from hedge funds.  These investors obviously want to see returns, and are sometimes taking direct action to get it.  Seth divided the hedge fund activists into three camps.  The first camp were the "alpha-activists," funds like Bulldog and Pirate Capital who use activism as part of their investment strategy.  They seek companies who are mismanaged (in their estimation), obtain a significant position of ownership, and then agitate for change.  The second camp are the "one or two-timers."  These are funds who don't typically get active as shareholders, but may do so if they see a problem AND when communicating their concerns to management feel that the managers have lied to them or not taken any remedial action.  In this circumstance they might go to a proxy fight to get a board seat.  The final camp are the "wolves," or opportunistic hedge funds who lie in the weeds and pounce on a company that is entering a proxy situation.  They pile in and, in sum, can obtain 30-40% of a company's shares - at which point the management team can not win a clean victory and must concede something. 

I have a feeling the farsighted and skilled hedge funds who have reaped outsized returns using these methods have signaled same to the market.  Now the entrants will be less skilled and more aggressive as more money chases fewer good deals.  The funds who sprung up a few years ago saying they would be "moneyball" funds or "long-short" funds or whatever the flavor of the day was will now rebrand themselves as "activist" funds and we will likely see a lot more proxy votes and perhaps a small return to the hostile takeover atmosphere of the 1980's. 

Of course, most of the latecomers will fail to get the returns they hope for and this too will burn itself out, leaving only the investors who actually know how to agitate properly.

One such fund is ValueAct Capital.  I believe their CEO, Jeff Ubben, when he says he takes a private-equity stance in the public markets.  He looks at his investments over years, and repeatedly talked about how important it was to win the "moral high ground" with the management team, who are understandably nervous about someone showing up on their doorstep owning 5% of their company.  Like soothing a skittish horse, it takes a lot of talk and slow movements to get to the point where the management team will listen to what the investor has to say let alone give them a board seat. 

One benefit of having the investor get a board seat is to have a close and real-time communication between the management team and someone who represents the shareholders.  As Jeff pointed out, "For years they've been guessing what the investors want.  When I'm on their board, all they have to do is ask!" 

But to "do it right" - the way I believe Value Act does it - a fund needs to have the proper capital structure.  You can not have money that expects outsized returns in 90 or 180 days.  You need to lock them up for a few years, but have a diversified enough portfolio of investments that a few growth engines will be firing while the newer investments take their time to mature. 

When the Obvious is Unobtrusive

Salesforce.com is a SFA application, right?  And it's hosted.  Oh, and they bill you over time.  These are the kinds of responses that even some savvy investors/IT pro's might give you if you asked them what they know about Salesforce.com.  While all three are correct, they miss the larger reality of what Salesforce.com really is.

It's a huge online database.  With a small application on top. 

Today at the analyst day at DreamForce, CEO Marc Benioff made this point repeatedly.  And to be sure, he has been making this point for as long as I've heard him talk to analysts.  But it still has a novel ring to it whenever I hear it. 

Marc's DNA is that of Oracle.  The dude knows databases inside and out.  He also grokked the role of the internet and, together with Ray Lane, spent many late night skull sessions talking about what an "internet database" might look like.  The basic question was around what had to be done to the DB to let it take advantage of the orders-of-magnitude scale the internet was sure to provide.  And now that bandwidth is not really an issue, and we have a universal client, why not try to offer a service where software is a utility that pops into your flatscreen via a browser from anywhere at anytime?

This gets to some important realities of what SaaS is and what is isn't.  On Demand is really a huge virtual database in the sky somewhere that has all your data and allows you to access it how you want in a way that gets a job done or otherwise provides value.  On Demand is not a traditional software application offered from a hosted system.  Although for the time being, it's OK to think about it that way.  We are in the midst of a middling paradigm shift and it will take some time to develop a thorough understanding.  But defining "On Demand" as a traditional software app that's hosted misses the point - the software isn't the value.  It's the database.

Oracle made their mark with the finest database on the market.  It was secure, stable, scalable, robust, and in most critical categories technically superior.  Think about Salesforce as Oracle, but native to the internet and you start to get the picture of where Marc is going, and the possibilities before them.

CRM Up Close

I spent the past three days in Chicago at the Gartner Group 2006 CRM Summit and came away with a few tidbits worth sharing.  In general, the Gartner analysts were bullish on the overall sector as companies are investing in CRM as part of their top-line growth strategy.  Specific technologies under the spotlight include SOA, data quality, and BPM.

CRM Growth: The theory put forward most frequently basically asserted that, since the bursting of the bubble, corporate America has spent the past five or six years hoarding cash and cutting costs.  While the economy seems to be in the midst of a prolonged period of indecision between recession and boom, we seem to be in a period of steady, incremental growth.  As such, corporate America is increasingly turning to a focus on growth.  From an IT standpoint, one of the key investments for growth is in CRM. 

Under the covers: CRM is treated as the sum of three sub-segments by Gartner.  The first is Salesforce Automation (SFA).  This is basically opportunity management and reporting.  Seibel and Salesforce.com are strong here, although the basic functionality is pretty well commoditized.  The second segment is Marketing.  This segment focuses on managing the results of marketing campaigns with a focus on email management, conversion of marketing lists to prospects, and reporting.  The final category is Customer Service and Support.  Think of call centers and the technology that supports them: call logs, screen pops, IVR and the like.  PeopleSoft CRM has strong roots here as does RightNow. 

What I found interesting:  While SFA holds the lead as the largest segment by revenue, it isn't much more than the total revenue of CSS.  Furthermore, Gartner predicts that next year, CSS will eclipse SFA for the top revenue segment of CRM. 

SOA: The underpinnings and infrastructure of these CRM applications are functionally similar.  The basic requirements are scalability, reliability, and security.  From there, IT shops are focused on resuability of components.  SOA's maturity and development of open standards enables a more agile, and reusable CRM application. 

The upshot: R&D in CRM overall will focus on SOA for the next year or so.  This means that the development of feature/function will slow. 

What I found interesting: Business customers don't buy SOA.  They hardly understand it and care only slightly less.  They buy tangibles, such as opportunity management, reporting, analytics, or campaign management.  Given this assumption, the players who are strong in those areas will continue to find success for the next year or so.  This doesn't diminish the importance of SOA in an evaluation - it appeals to the IT buyer.  And a good rule of thumb to remember is "IT can't say 'yes,' but they can definitely say 'no' to a deal."

Data Quality: With the advances in technology such as SOA, the silos of information are giving way and creating the opportunity to leverage ALL of a firm's data.  The trick is to ensure that there is "one source of truth."  A survey of hands showed that most companies in attendance had 60 customer databases.  Deciding what data in which database are the best at describing the customer, for example, is a challenging undertaking. 

The upshot: A cottage industry of data management firms have popped up who specialize in this.  Siperian, Purisma, DataFlux, Cordys... These complement the big boys of SAP, Seibel, and IBM who also have offerings in the space. 

What I found interesting: The beauty of this will be apparent to all when you interact with a vendor and every touchpoint - call center, email, direct salesperson - has the exact same information about who you are and you don't have to constantly re-educate them in order to get anything done. 

BPM: Once the underpinnings are solid, a company can lower costs and perhaps develop a competitive edge via their processes.  A good BPM tool can enable the design of these processes. 

The Upshot:  A more agile, nimble IT that can adapt to the market. 

What I found interesting: Not much, really.  BPM is apple-pie and motherhood stuff.  Gartner is right to focus on it because the payoff is huge, but I feel that if you AREN'T focused on your processes then you probably won't be around long anyway. 

The Daily Dish, HP-Style

So this article is a nice summary of some goings-on at the HP.  In a nutshell, Mark Hurd was fed up with leaks that originated at the Board level and took some action that may or may not have been legal.  In response to this, Thomas Perkins resigned.  And by "Perkins" I mean THAT Perkins - of Kleiner Perkins fame.  A technology celebrity catfight is officially on!

The article does a much better job than I could relating the facts without speculating on motives behind the actions.  That is a crying shame, and to fill the vacuum I will insert totally unsubstantiated speculation as to why certain things happened.  Do not attempt this at home, I am a professional idle speculator. 

Fact one: Mark Hurd was fed up with Board Level Leaks. 
My take: Of course he was!  It's tough to present an image of a unified board when certain of them are whispering to the press or Wall St. something other than what the CEO and Chairman is saying.  Authority undermined.  Ability to govern questioned.  A separate center of power develops that attracts your enemies.  None of this is good for your longevity. 

Add to the fact that Hurd's predecessor, Carly, was similarly incensed about Board Level leaks.  She had her hands full with the Compaq merger (which at the time I gleefully called "one midget standing on another midget's shoulders exclaiming, 'look how tall I am!') and was not well served by her leaky board.  Hurd had to act to seal the leak, or risk being undercut on every decision.

Fact 2: the "pretexting."  Apparently the agency hired by Hurd to smoke out the leaker used something called "pretexting."  This is a practice that has someone call in to, say a phone company and pretend to be someone else.  They don't give confidential information but simply try to outwit the call center rep on the other end of the phone to obtain confidential things like phone records.  Which they did. 
My take:  Wow!  Lawyers for the agency say that no laws were broken but they aren't really sure.  As a guy who majored in Philosophy and took more than his share of Ethics, let me tell you "pretexting" as a general practice fails almost all major ethical standards - Aristotelian, Kantian, Consequential, Utilitarian, and Platonic.  But in this case strong counter arguments could be made if you think that Hurd owed it to the shareholders, employees, and customers of HP to stop the leaking.  If you grant him this standard, I can see arguments that keep "pretexting" from violating Utilitarian but still failing Kantian and Consequential.  Aristotelian and Platonic are up for grabs - does this appear to be virtuous?  Do you care that it has been made public?  I can see arguments on both sides.  So maybe it wasn't that unethical.  But it sure feels slimy. 

Fact 3: Thomas Perkins resigned and cited the process for outing the leaker as the reason why.
My take: I call bullshit.  Thos Perkins resigned over a new CEO sealing up leaks that absolutely should not have happened?  No way, unless he's not compos mentis.  So what's in it for him?  Loyalty to the leaker?  Possibly but that's not enough.  What about the chance he was about to get outed and he simply beat them to the punch?  Possibly but the article doesn't suggest that at all, and after all they (presumably) did "pretext" his phone records too and didn't finger him.  I'm guessing it was a combination of things - a new CEO who he didn't like in the first place was moving to eliminate a strong ally of his, thereby reducing his power in the press, company, and financial communities.  (Remember, information is king.)  I think there has to be another element to this - maybe he was just tired of HP or perhaps they DID have some dirt on him.  Something else, but all together creates in my mind a reason for a public outing of Hurd as a guy willing to go to great lengths to do what he feels he has to do, regardless of the law or ethics. 

Fact 4: the leaker was George Keyworth. 
My take: Huh?  Who?  I had to read a few bios to conclude the guy looks like a stud.  Scientist, politico, connected...  But why screw around where you shouldn't?  Perhaps he confused his role with that of the CEO?  The motives for leaking are diverse - and made more complex by the fact he was leaking information that didn't exactly help the stock.  I will assume he was (is) a megalomaniac who liked accreting power. Safe bet!

So what do you think?  Comments welcome. 

BEA Makes Insightful Acquisition

Yesterday, BEA Systems acquired Cleveland, OH based Flashline Systems.  As the saying goes, "actions speak louder than words" and this move suggests BEA actually has real-world SOA experience. 

Flashline makes a meta-data repository for SOA objects.  To be sure, there are registries of information (think UDDI) that give the address and location of a web service.  But when building SOA applications at design time, these registries fail to contain important data about the service that directly impacts the system design.

For example, it isn't enough to know where a certain pricing engine resides, a designer really needs to know what it does, who has permission to use it, what are the transaction limitations (if any), and so on.  Depending on the module, there could be many more such important pieces of data and Flashline solves the problem of managing that information. 

Ultimately, SOA's ROI is in reuse of components.  By providing more insight on the inventory of services, a designer theoretically can create better systems faster.  And isn't that what infrastructure is all about? 

By this acquisition, BEA tips off that they are working on many SOA projects and are running into this SOA design problem.  Kind of an advanced problem, and not one that I believe is in the mainstream. 

Obviously the value of this move depends on the price, which was not disclosed.  But I think this solidifies BEA as a heavyweight in the rapidly expanding SOA battlefield. 

A Fifth Open Source Hurdle

Can't believe I missed this one.  And I've even interviewed Darl McBride!

#5 Patents.  Most open source projects are just that: projects.  They are designed to improve something that already existed in the OSS world or create something their business needs demanded.  The result is lightweight and expeditious and made freely available. 

In other words, the project has no money.  And with the proliferation of patents in software, this means a given project has no money to defend itself against a patent infringement claim. 

A massive proliferation of patents and the subsequent vigorous enforcement would put a serious dent in the OSS movement.  MMicrosoft, for one, has indicated it will become much more active in their patenting of their programs.

This article talks about how the EFF is trying to persuade the Justice Department to do away with the "Suggestion Test."  This is a standard "the U.S. Court of Appeals for the Federal Circuit has used to determine whether an invention should be obvious to others, and therefore, not patentable."

Apparently the onus is on the DOJ to search through mountains of published data to see if someone has made a reference to the obvious need for a program.  If this is the case, the patent is denied.  If no reference is found, the patent is not considered obvious and goes on to the next step.  The EFF is arguing that this is really a measure of the DOJ's ability to sift through an incredible pile of data to find hard evidence.  They further point out that this method doesn't really address the "obvious-ness" of an idea - sometimes it's just... obvious.  If it's that obvious, people won't necessarily write about it. 

Of course, the EFF has a dog in the fight.  They want to make it much harder to make software patent claims so the Free Open Source Software movement can exist relatively unimpeded by patent claims. 

Be interesting to watch how this ruling goes.  I tend to agree that the current system for granting patents needs to catch up with the Information Age.  Amazon's patent for "One Click" ordering?  Are you kidding?  They INVENTED that?  And it WASN'T OBVIOUS?  Had ANYONE at the DOJ ordered anything online when that patent came out? 

Four Open Source Problems

This is a well researched and written article that can only scratch the surface of a major challenge facing the maturation of open source.  Fundamentally, open source is in their teen years - lots of promise but still immature.  It's a mess of overgrown limbs and a brain not yet used to how they function. 

I see at least four hurdles that need to be cleared before Open Source ascends to where it should be - a credible alternative to "traditional" software in all aspects of the concept.

#1 - Licensing.  There are so many licensing options out there it hurts the head to read them all and figure out the implications to your product or service.  Right now, the only winners in the current licensing fiasco are the lawyers.

#2 - Monetization.  The "download-try-buy" idea of early open source is great.  It means your product has to be brain-dead easy to install and use.  It means you only get paid when someone finds value in it.  It means your sales and marketing costs are low.  But recently, I've noticed the "free" version of a commercial open source vendor's product is so hamstrung that it's almost useless.  Such vendors are traditional software vendors in disguise - they have per user pricing, per CPU pricing, the whole shebang.  The "open source" version is nominal at best and they protect their IP with custom licensing while charging traditional prices for it.  Hmm.  That isn't much progress.

#3 - The blend of "core" and "community" developers.  The open source community has many different models of development.  Some have most of their code developed by the core team, and then they flip it out to the community for use.  These community developers use it and feedback bugs and suggestions for improvement.  Some actually can submit new code for use in future versions, but the core guys have the control and final say of what is supported.  On the other end of the spectrum is a huge community and weak core.  While this model takes advantage of the communal intelligence and delivers high quality software, it also is rife with forks that cause support headaches.  The correct blend is not clear yet, and the trade-offs need to become clearer to the mass market.

#4 Trust - Who is certifying the open source stack?  Well, SourceLabs and SpikeSource are trying to.  But given the different number of licensing options, the  development models, the forking, and the monetization schemes, there is a problem of scope when it comes to certifying or supporting an open source stack.  It will take a lot of time and money to develop a credible vendor of trust who can own the trust in the infrastructure which is critical to support the applications et cetera. 

Keep an eye on how these four get rationalized. 

Also: what did I miss?

There is a Disturbance in the Force

Computerworld published an article a week or so ago entitled "Salesforce.com Bites Bullet on Need for More Integration."  Cited in the article are some revealing quotes that point up a challenge facing companies like Salesforce:

"If I were to say one single reason for our change, it was that we needed tight integration," said Flora Sun, chief operating officer at Adina for Life Inc. in San Francisco. "We needed fully functional end-to-end processes."
-and-
Portland, Ore.-based Learning.com cut over from Salesforce.com to NetSuite last September in order to get an accounting system that was tied to CRM applications, said Douglas Stein, vice president of technology and development at the provider of online educational services.

This is not a huge negative for Salesforce, but it does illustrate one of the main challenges of the shiny future world comprised of composite applications stretched over a global network.  The forces impelling SaaS are strong enough to provide the incentive for companies like Salesforce to solve this problem by creating more custom-built connectors, and to see expansion in the standardization of integration. 

One potential winner is NetSuite, Inc.  They provide a suite of services that sidesteps the integration problem as long as you use all their services.  That alone might be a tall order for large enterprises, but should be a compelling value proposition to some decent-sized middle market businesses.  Of course, the other trade off is the fact that NetSuite will lag category leaders simply because they do not have the bandwidth to be best-of-breed across multiple categories. 

Ultimately, these integration problems will be solved.  The traditional way to buy and consume software is too rigid, cumbersome, and expensive.  Each of those realities drives the need for alternatives like SaaS.  But let's not get too rosy - there are always growing pains. 

Today's Rant: Software Self-Reporting is Fun!

Recently, Oracle claimed their middleware business had reached one billion in revenue.  Wow, you might say.  That's pretty big.  And you would be right - a billion dollars in revenue is an important milestone.  It means you've been successful horizontally, vertically, and globally.  The market has spoken, and they like what they've seen.  The trouble is, unless you are a single product company - those numbers are easily manipulated and without oversight.  In other words, they're useless.  Actually worse than useless... they're mendacious. 

Here's what happens when you are a single product company:  your salesforce sells your product and revenue gets recognized.  $1 of license revenue means $1 of product sold.  $1 of maintenance equals $1 of service on that software. 

In a multiproduct company (think Oracle, IBM, CA, etc), the situation is decidely different.  Your salesforce may sell a single product and the revenue reporting is exactly as in a single-product company.  But your salesfoce may also sell a multiproduct contract for a fixed fee.  This is oftentimes called and "ELA" or "Enterprise License Agreement."  ELA's cover many products and sets specific quantities of those products.  The ELA is usually driven by the vendor's core product and in exchange for deep discounts the vendor will toss in some low-share products that they want to grow.

For example: A BEA rep may have driven a million dollar WebLogic deal, and then thrown in some Portal, Integration, and Security.  An Oracle rep might have sold a ton of database and then thrown in some app server, some BPM, and some integration.  You get the picture, but here's where it gets really ridiculous.  All you have to include is one single license of the "other" software to enter the unregulated land of carving up such a multiproduct contract.  80% of the value might be the core product, but with no oversight it's easy to re-characterize the composition of the contract to be 25% of the core product and 75% of the product the customer didn't need/want.  In the BEA example, what might have been 80% WLS would get recharacterized as 25% WLS and say 75% WLI.  In the Oracle example, 80% might have been RAC and after recharacterization that number falls to 30% and the app server gets 70% revenue credit. 

There is zero transparency and next to no oversight to this process.  To be fair to the software vendors, there is a sense of "don't lie too egregiously" but the standard of when to stop lying is pretty vague and along the lines of "at least keep it in the ballpark of reality."  So you don't see out-and-out falsification, but it is pretty bad. 

Of course, all vendors have the true numbers in their BI systems, but they cough and mumble when analyst ask them for specifics.  The usual line is "sorry, we can't break those numbers out" or my favorite baldface lie: "I don't have those numbers offhand right now." 

But who can blame them?  Vendors are not required to break things out by the SKU and there is no upside for being the only honest guy at the liar's ball.  At the same time there is a ton of downside to their stock price if a future growth product was actually languishing.  Wall Street's hypocrisy is part of the problem - they want the truth but won't reward the honest firms. 

But wait there's more.  The salesforce is incented, for lack of a better word, to lie. 

When a company has a new product that is supposed to be the next engine of growth, naturally a salesplan is created that gives a better commission rate (or some other financial reward) for selling that product.  When the sales rep closes their ELA and is asked to divvy up the pie for reporting purposes and there is virutally no oversight, it is no surprise that the higher commissioned products magically take up most of the pie.  So that BPM tool that really sucks but has a better commission rate suddenly is 50% of the million dollar deal.   Good lord, get the press release out!!!

And divvying up these complex ELA's befuddles industry watchers like Gartner/IDC.  They have to go by the data given them by the vendor, so how can they really know?  Unless there is legal oversight to the process, take all claims with a huge grain of salt.  The truest gauge of a product's competitive standing is how many mission-critical apps does it support?  Until there is a survey that measures that, we're all left in the dark.

A cynical smile gets drawn on my face whenever I read headlines like these:

• BEA WLI grows at 140% year over year - really?  We'll take your word that your integration offering has 140% more revenue than some other fictitious number from a year ago.  Bravo!
• Oracle Middleware Reaches $1B - so you're telling me that you went from zero middleware revenue to a billion from 2000-2006... a span of time that included the two worst IT spending years in history, and with no middleware products that anyone considers first class?  (Note: I consider database it's own market and category... obviously Oracle excels there!)
  
• WebSphere Makes Major Marketshare Gains - Wow!  IBM must have beaten BEA head to head thousands of times!  Or they threw in some WAS with every sale they made globally. 

Ah software.  We luv ya!  It'll never change, and so we are assured a limitless supply of ridiculous metrics that claim decisive victory.  Just don't believe any of it.  My fundamental rule is: If there is no government oversight or natural market advantage for honesty, then assume it's a lie.  Put the burden of proof on the firm to prove their claims and go from there. 

The Fourth Business Unit

Traditional software vendors (read: On-Premise and Perpetual License model) are struggling to grow.  The era of hyper-growth for these companies is in the past and the present finds them engaging in a battle for mind/wallet/market-share.  This article in Computerworld details an overlooked aspect of the software industry: the audit.

Money quote:

Most IT budgets are fairly flat, and the only way the industry is going to survive, in their opinion, is by increasing wallet share. One way of doing that is by auditing and using that as a mechanism to generate revenue.

Increasingly, companies squeezed for license revenue are turning to 3rd party auditors to survey their installed base and ensure that nobody has over-deployed licenses.  A software firm's right to do this is a standard term in any license agreement.  The reality is that software agreements are oftentimes good faith agreements to deploy X number of CPU's or Y number of users.  Since there is no effective real-time way to measure this, a market for auditors has been created. 

The auditor, working with the CIO department (who is obligated to assist the inspection or face legal action), can determine the number of CPU's or users and compare that to the original deal.  If a company has overdeployed, which is a frequent case, the customer has a period of time to get back into compliance or pay consideration to the software vendor. 

These audits have become so successful that one company (that shall remain nameless) refers to their auditing team as "the fourth business unit." 

Not exactly romantic, and not something that I reckon is sustainable as a revenue strategy.  Once a company is stung with an out-of-compliance bill, I guarantee they improve their internal processes and controls such that future mistakes along these lines are minimized or eliminated.  Not only does the financial surprise irritate customers, but the distraction the audit causes also cuts into vendor-customer goodwill.  In a nutshell, this pisses off customers. 

Beware of investing in software firms whose license lines are buoyed by these tactics.

Hurd opens fire

The deal is finally done.  In what made a ton of sense for both parties, Hewlett Packard acquired Mercury Interactive today for $4.5B.  I like this deal for the most part, although I was not completely sold on the idea that Mercury *HAD* to sell to Hewlett Packard.

Mark Hurd buys instant credibility in the testing and governance space while augmenting their systems management product line.  Mercury's products received high marks from customers in general for doing what they were advertised to do.  Of course, Kintana at first was more promises than anything and Topaz still requires too much PS, but in an industry that over-promises and under-delivers, Mercury was a cut above. 

The other huge asset HP may retain are the Mercury people.  These folks were talented and motivated.  If HP can retain key executives, sales, and staff, HP looks to have made a good purchase.  This is of course, always a huge "if."  But I'd look for future HP leadership to come from some retained Mercury exec's, similar to how Shai Agassi became a leader at SAP via TopTier. 

When Mercury was dinged by the stock option fraud, the market naturally corrected to reflect the uncertaintly that there may be another shoe to drop.  After all, Amnon Landon was regarded by some as one of the white-hats and above reproach.  When he went down, I think some investors lost a ton of faith in software overall and especially in the MERQ name.  Some savvy investors that realized the company was fundamentally unchanged bought a ton of stock at 36 and 37.  Those same investors are smiling broadly now. 

The psychology on the Street soured so much on MERQ that an acquisition almost became fait accompli.  It amazes me that once enough of the Street decides you should be bought, it's almost always safer to just sell out because turing that tide of opinion takes years.  Quarterly successes are met with indifference and any mistakes are punished excessively. 

Despite this practical consideration, I thought that MERQ could soldier on solo.  They had the products, the strategy, and the people.  The market was moving into a somewhat turbulent stage with the advancement of SOA - this was a challenge and an opportunity for a company like MERQ.  But it was pretty clear that they had struck the colors the moment Amnon was forced to resign.  Adios Merq, Hola HP. 

Web 2.0 in the Enterprise

This week found me in sunny San Diego at the Gartner Group Application Integration conference.  The key takeaway is the state of Web 2.0 in the enterprise. To be sure, Web 2.0 is transforming (and being transformed by) the consumer space. People today blur the lines between work and home life; they expect to be able to access their work tools from home, and their iTunes from work. People today generally have a much greater depth of computer skills and have internalized the web to their lives. Popular consumer software includes the aforementioned iTunes and IM clients. Downloading plug-in’s to one’s browser is not a tall order nowadays. It’s pretty simple – download, install, setup, and start using. No manuals, no phone calls. Where is this in the enterprise? Where is the leadership coming from? What might a commercial Web 2.0 company look like?

AJAX, RSS, POX, and REST are the four acronyms of heralding the brave new world of Web 2.0. This version of the web behaves more like a client application while is constantly refreshed in a personalized organic way. Leadership is coming from the most unexpected places – Zillow.com did not come from Microsoft or IBM. Del.ici.ous, Flickr, and HousingMaps were created by unknowns.

Please don’t bore me with the “this is nothing new” argument or the “where’s the revenue model” observation. Nothing of itself is particularly new, but the confluence of all this new-oldness combined with broadband and critical mass in interoperability most certainly is new. The revenue in Web 2.0 is still being hashed out but advertisements seem to be working for now, and I am sure someone will square the circle for micropayments down the road.

But Web 2.0 transcends technology. It enables human behavior in another medium. Social networking and communal knowledge with peer review are not new but on the internet are startling – for now. Beyond the voyeurism of reading personal blogs or the amazing facility of Wikipedia lies a
Eureka moment in the enterprise. In 18 months or less there will be a story in Wired about the equivalent of MySpace.com for the enterprise. In other words, there will be a killer Web 2.0 application – for the enterprise. And to confound you even a little more, this application might not be commercial. 

It will probably be stitched together with an open source application server and tools, where services are created and exposed for reuse. Perhaps someone in another division finds these services and mashes them together with some data in their department. A business manager sees this and the lightbulb explodes. Maybe it’s as simple as crawling around the internet and identifying all the blog posts about your product, categorizing them by tag, and feeding them into an analytic program to observe near time feedback trends of your product?

But between Web 2.0-in-the-enterprise-nirvana are some serious obstacles. And just as Web 2.0 enables positive human behavior on the web, so too must it deal with the impediments of normal human behavior. For example, enterprise culture matters a great deal. If your company does not have a culture of sharing and trust, then it is highly unlikely you will be leading the charge into department social networks and grass-roots knowledge management. There will be no wikis or blogs to share opinion, fact, or other data.

Habits are also hard to break. I heard developers, thrilled with the possibilities of Web 2.0 in their enterprise, lament the fact that nobody on the business side asks them to expose an XML api for reuse, or to use POX, or learn AJAX. I found it hard to keep a straight face when this comment was sincerely made. The business side has been conditioned to ask only for incremental features and not out-of-the-box ways to collaborate or leverage the tremendous amount of data out there. The IT guys are tired of impossible and fantastic requests from the business side, and they have had to severely regulate the creativity coming over from the business side.   

But in this inefficiency lies the next breakout hit. The company with the culture of sharing and trust, and has a decent level of IT savvy with the business users combined with a modicum of business acumen from the IT folks will suddenly be the next WalMart. They will create dynamic extranets with their top suppliers, partners, and customers who share the same vision and values.  They will generate loyalty and referral business. 

A hallmark of Web 2.0 is “good enough.” That is, write software that gets the basic job done, post it as beta, and let users start slamming on it. The software vendors must then capture the feedback and rapidly release improvements requested from the market. However, in a revenue-generating e-Commerce application “good enough” isn’t good enough.

Or is it? Much like the hellish TPS reports from the movie Office Space, many companies foist meaningless work on their people that are designed to improve operational efficiency. While this is a noble goal in a stable market, if Web 2.0 is anything like Web 1.0 (and it will be) this is not a stable market. Firms who punish a dropped transaction so severely that it impedes the creation of an out-of-the-box way to approach a market may be on the losing end of the game.

For example, an e-Business manager at a well-known bank has breakdowns by the hour of new client adds. If the bank fails to add a client due to downtime, this manager gets the business end of a hot skewer. This manager therefore runs his business to five nines uptime. This manager is also in no position to dabble with Web 2.0 combinations with his knowledge base, offline networks, collective intelligence, etc.

Am I advocating the loss of a few client adds (and corresponding revenue or bad word of mouth) in the favor of tinkering. Um, kinda. I would argue that the reward given to the first companies that get it right is so great, so jaw-dropping that now is the time to spill the milk, break the dish, and worry about the maid service later. The companies that worry about their TPS reports will be reading in the papers about the valuations of the companies that went for it and got it right.

Of course, I am insane. Long ago I realized I am highly risk tolerant. It wasn’t until grad school that I saw studies that proved, statistically, most people will not take a bet where they have less than a 3-1 chance of winning. I learned I would take a 1-1 bet all day long, just to see what happens. I am also in the extreme minority (~1%). So yes, I will concede that most people who neglect their TPS reports in favor of learning AJAX will get burnt. Knowing this need is there, it is a matter of time before a commercial entity appears that can take the well-intentioned but poorly-led/skilled/whatever.  This entity might have an AJAX developers tool that automatically generates AJAX code from a visual-studio paradigm. They might have a set of Web 2.0 examples that fire the imagination and introduce a whole new level of innovative ideas. They might also have the mentoring, training, and program management to execute a plan once defined. They could do staff-augmentation.

Tibco may have been thinking about this when they acquired General Interface, an AJAX development tool company. Another out there is JackBe. And one that takes Web 2.0 as a concept and stretches it farther than anyone I’ve met with is denodo technologies. I hope I have time to blog more on them in the future. Check them out!

Web 2.0 in the enterprise has some serious hurdles to overcome, but the payoffs are there. It will happen, and the leadership will come from strange new places. The revenue model will likely be advertising-based, and the lexicon of new successful software companies will expand by a few. Describing it today is like trying to understand the Fifth Dimension. Analysis suggests strongly that it’s there, but nobody can explain it adequately, and we’ll only know it when it shows up on the front page. By the same token if the killer Web 2.0 application for the enterprise comes screaming across the mediasphere and you aren’t involved, you will miss the second big revolution.

Arms Merchant for the Next Big War

One of the great things about living in San Francisco is the proximity to the tech startup community.  It's pure, uncut optimism. 

A few days ago I was fortunate enough to visit with a very hot startup called Zimbra, located in San Mateo.  It is being run by the former CTO of BEA, Scott Dietzen. 

Zimbra is Web 2.0 Exchange.  It's all Ajax, it can be delivered either on-premise or off-premise, its open api's enable some eye-popping mashups, and it leverages an open source development model.  What is cool is that Zimbra was developed after the creation of the Web which means it can take advantage of many nifty standards, technologies and techniques.  Some would say this creates a "second mover advantage."  Why?  Here's a shocking stat: Exchange's code base is 10 years old.  This means that the genesis of Exchange included requirements to handle things like PROFS, a mainframe based email application. Exchange was built before many important web-based languages and concepts existed.  In a nutshell, Exchange was built in 1996 with one eye on serving their client/server requirements while also anticipating what the web was going to do to messaging and collaboration - an impossible task.   

Zimbra, by contrast, was developed with web concepts at top of mind.  Careful consideration was given to how Web 2.0 enhances things like sharing.  It is far simpler to have multiple personal calendars - one for work, one for personal, one for your avocation... and then share them with different constituents.  I know something like this is possible in Exchange but in Zimbra it is a breeze by comparison to set up and right now appears more powerful.   

Zimbra lets you do eye-poppingly cool things with emails and contacts that should have been features in Exchange ages ago.  For example, hover your mouse over an address and a map pops up.  Or hover over a flight number and see that flight being tracked by www.flighttracker.com.  And that's the heart of what Zimbra is - it is a much more flexible, simpler, cooler, and better looking Exchange.  However, isn't this strength a weakness?

I will not argue that Zimbra is cooler than any other mail package out there.  It is.  I want it for myself.  But can't all this coolness find its way into Exchange?  I am not a technical guy, but the answer I gleaned was along the lines that Exchange probably could do it, but it's much less elegant and inflexible.  Hmm. 

And then you realize that is not the point.  Zimbra is not out there to unseat Exchange.  It is out there to provide the same basic services in a much more natural, intuitive fashion... priced as a service.  While they are still in the early stages of their distribution strategy, I found it very interesting to notice one of their strategies is to partner with telcos.

Suddenly I knew why Microsoft should care, and why Google will care.  With Microsoft launching Office Live (a hosted set of services for SMB's, including email), they positioned themselves to capture the customer at the point of conception.   Objecting to this move, interestingly, are the telco's and ISP's.  The real dollars lie in the services delivered through the pipes of the telco, and they will be God-damned if they let Microsoft saunter in and capture that revenue without a fight.

The arms merchant to this potentially huge battle then becomes, among others, Zimbra.  Now British Telecom or Verizon can not only offer the bandwidth but also a compelling, easy to use service that will win heart and mindshare from anyone on the other end of the pipe.  At the very least, it is legitimate competition to Microsoft.

Considered in this light, Zimbra becomes quite interesting.  And one would think that Google, with their apparent strategy to offer a web-based version of all things Microsoft, would take a long look at Zimbra.  Competition for this space include Scalix, PostPath, and Semaphor.

In the mean time, I imagine the gang at Zimbra will focus on fleshing out the core feature set of the product, and enhance the collaboration/sharing powers.  If they can convince the large, medium, and small network plays to resell their stuff, we'll have another success story, and another reason to be optimistic about Silicon Valley.

Global Integration Summit

This past week found me in Boston for the Global Integration Summit.  This consortium bills itself as vendor neutral and specializes in "establishing standards, guidelines, best practices, research and the articulation of strategic and measurable business benefits."  The theme of the event was "Integration for Everyone" and was attended by about 125 people at the Fairmont Copley Plaza. 

The event reinforced what I had been hearing and thinking about SOA.  It is for real and it's one of three significant trends in software today (the other two being Open Source and SaaS).  Conversations between participants were well beyond skepticism and focused more on what this meant for their business.  SOA has reached an important milestone - call it 1.0 - and is out of the bleeding-edge laboratories and into the early adopter phase of existence.  There are some important kinks to be worked out before we can say it has jumped to the next level.  Once that happens, SOA will be solidly in the early majority phase and will rapidly enter late majority as some higher-level standards and concepts mature.  All up, this final phase is 3-5 years out. 

SOA
Stop me if you've heard this one before: there's a new architecture out there with new acronyms and religious zealots proclaiming "this time it's different!"  Now you can make the smug observation that SOA as a concept is about as fresh as a Foghat concert.  I am aware that as far back as 1972 David Parnas was writing a book called "On the Criterion for Decomposing Systems into Modules."  The whole notion of creating an infrastructure in an a priori fashion has been in the "Holy Grail" category for a long time - basically building the chassis and frame and plumbing so that you don't have to think about what will be built out of/around/on it.   

One of my favorite Wall Street analysts, John DiFucci at Bear Stearns is a skeptic.  John is right to wonder why "this time it's different" given the long and unsuccessful history of SOA architectures to date.  But the reason I believe this will reach a broad base of customers is my belief that the key underpinning technologies have matured and are open.  And most crucially, there is a virtuous cycle ensuring that no single vendor will own those standards. 

I'm talking about XML, SOAP, HTTP, WSDL, and UDDI.  We have passed a point where consensus was created by dialogue and market forces.  Now these standards are so prevalent that anyone trying to write to a proprietary or alternative protocol will find their service of low-to-negative value outside of their firewalls.  In other words, publishing a service using these protocols enables reuse on a massive scale not previously seen and reinforced by the recursive network effect of using the protocols themselves.  When self-interest cements a standard, I will begin to consider it durable, and I am considering these to be durable. 

But these standards are not enough alone.  As one consultant at the conference put it, "we're at 1.0 of SOA.  In order to get to the next level, we need reliability, scale, and performance."  He pointed out today that there are standards like WS-Reliability and WS-Security that are being baked today, but until these are battle tested SOA will still be in the early-adopter camp.  It simply wouldn't be prudent to put mission critical apps on something that is not enterprise class right now. 

This is not to say that SOA is dead.  Far from it.  In fact, if companies of a certain size and scale are not dabbling with SOA right now they are behind the curve.  Right now is the time to be launching SOA initiatives to gather the in-house know how that will shorten the time to productivity when SOA adds more meat to the bones it has today. 

Ultimately, SOA will leverage things like BPEL and WS-Coordinate to dynamically assemble apps at runtime.  One additional challenge for the SOA crew is to work on performance.  Composite apps are dogged by the perception or reality that they aren't as fast as other apps.  XML is a verbose structure, and parsing it will be key to increasing the throughput of composite apps. 

I'm optimistic that all these challenges will be met because it dramatically reduces the cost of IT.  SOA allows almost any vendor's service to be snapped in or out, and lets the work get done wherever it's done most efficiently and effectively.  Furthermore, the a priori nature of SOA promises to remove a huge barrier to business agility - the inflexibility of IT systems to adapt to market needs.  When you have drivers such as cost reduction and business agility, it's time to pay attention.

A word of warning: the plateau of productivity is still 3-5 years away.  This is not a revolution, it's an accelerating evolution.  It does not have a drop-dead date like Y2K, but there is a sense of urgency in IT shops.  Overall, this is something worth investing in, but do not confuse it with something like Y2K.

Power to the Datacenter!

Not too long ago I trekked to the Santa Clara Hilton for Sun Microsystem's annual marketing push.  This year it was called "Powering the Participation Age" and I went, curious how they were going to push Servers on me. 

I sat in the back, per usual, thinking that there was about an 80% chance I would decide this was irrelevant to the worlds of software and capital markets.  I sat next to two guys from a major consulting/outsourcing firm who had driven 3+ hours from Sacramento.  I chalked them up as Sun zealots and was cynically bemused that Sun still had zealots. 

The pitch kicked off and I don't recall much standing out - I did hear about the open sourcing of Solaris, of Open Office, and the fact they open sourced the chip design for some of their SPARC chips  - until they began to talk about their Niagara chip line.  The gents next to me leaned forward, as did most others in the room.  The presenter made some claims that I didn't write down but were impressive.  Something along the lines of: "this chip runs at 70 watts, and performs better than our SPARC chips, and it's 1/10th the price." 

At the break I began talking to my neighbors from Sacramento.  As I suspected, they were there for the hardware, but what I didn't realize was their urgency.  They were in charge of the huge datacenter that was the backbone and lifeblood for their enormous (and I mean ENORMOUS) outsourcing practice.  They said their #1 priority was getting control of the costs it now takes to power their facility. 

What I learned next bespeaks how long it's been since I've been in the hardware world.  Evidently, as servers have grown smaller and more powerful, the power requirements have skyrocketed.  They told me that if your datacenter was more than 5 years old, it probably was obsolete.  Basically, the cost to open the pipe for more power was more expensive than just buying or building a new datacenter.  Furthermore, the power to run the datacenter created huge cooling costs that were similarly out of control.  I was agog. 

So these fine gentlemen from Sacramento were there to learn about Sun's efficient Niagara chip.  They had already purchased them and deployed them on the fringes of their datacenter for file/print kinds of things.  They were gearing up to stress test the chips and determine if they could offload some of their hotter, power hungry chips to these new chips.

I asked about competition, and they said the major vendors (Intel, Dell, IBM, HP specifically) had nothing on the horizon.  I did not ask about AMD and should have. 

So in summary, one of the flagship outsourcers in the world has purchased and deployed these new sun servers almost sight-unseen due to their massive power and heat problems.  Sun, to that point, had not spent one dime on a direct salesperson to assist in what they'd purchased. 

Needless to say, I was impressed.  As a (former) salesguy, it is extremely rare to see a customer with such a clear compelling event for purchasing product.  These guys looked like they'd been shelled in a bunker for a year and were desperate.  Whomever has a power-efficient (without sacrificing performance) server will have a chance to reap excess profits for a time. 

I wonder if software will also feel the pinch.  Will CIO's begin to look at all the bloat and overshot feature sets that require such power hungry servers?  Will that provide impetus to the "good enough" versions of software needs?  Keep an eye on it.  And keep an eye on Sun.

S&P Disagrees With Me

Forbes reports that S&P analysts Gary McDaniel and Steve Kessler find Microsoft "notably undervalued" based on anticipated earnings from online advertising. 

I haven't seen their model but my gut says they are quite optimistic about MSN behaving like Google in the near term.  My empirical senses did notice they cut their price target from $33 to $31. 

Put your money to work elsewhere.  Try BEAS, TIBX, WEBM for SOA plays. 

Microsoft the Stock

Microsoft is the dark center of the software universe.  By any metric, they dominate.  And despite being threatened by open source, litigation, and a bevy of new competitors, they will continue to print money for the forseeable future. 

The Windows operating system and the Office suite are a symbiotic monopoly that Microsoft leveraged to destroy all comers with only a few exceptions (attaboy Adobe!).  I will not argue that Microsoft won a fair fight on a balanced playing field because that is pure fiction.  They had a compelling market advantage and they did what nobody else could do.    

For the stock, this has huge ramifications.  Even today, Microsoft is fundamentally just a desktop software monopoly with more cash than they know what to do with.  Anyone investing in this name is essentially investing in the growth of PC shipments and/or the expectation of a dividend.

Microsoft's latest quarterly income and operating margin numbers:

• Client (Windows) : $3.187B in revenue, $2.478B of operating income for an operating margin of 77.8%
• Information Worker (Office): $2.946B in revenue, $2.087B of operating income, 70.8% operating margin

Pretty nice to have a monopoly, eh?  For those of you scoring at home, companies who actually compete for business are thrilled to find operating margins north of 20%.  Most are in the low teens. 

So now the rest of Microsoft's business.  These are the business units that actually compete against real rivals and can not leverage the desktop monopoly as effectively.  Although they do try.

• Server and Tools (SQL Database): $2.845B revenue, $1.068 income, 37.5% operating margin
• Dynamics (ERP Solutions): $216M revenue, $-13M income, -6% operating margin
• MSN: $561M revenue, $-26 income, -4.6% operating margin
• Mobile and Embedded Devices (Windows CE): $89M revenue, $-14M income, -15.7% operating margin
• Home and Entertainment (XBox): $1.056B revenue, $-388M income, -36.7% operating margin

The competitive businesses sum to $4.767B of revenue, $627M of income, and 13.2% operating margin.  That is still a staggering quarterly revenue number, and you have to call the Server and Tools business a success, but every other line of business is funded by the Monopoly.

I understand the argument that these are long term investments.  But who really thinks they will ever pay off on any sort of scale?  Every one of those "investments" has significant and well funded competitors who will vigorously defend their turf. MSN is in 3rd place behind Yahoo and Google, and has had enough time to make more of an impact.  Hard to qualify MSN as a success.  Dynamics is a cluster-smack of four different product lines (Great Plains, Axapta, Solomon, and Navision) that still operate like four different companies.  They are focused on the sub- $1B accounts and give the larger deals to their parter SAP.  Nevertheless, the going is very tough against the SMB offerings from Sage, JD Edwards, PeopleSoft, et cetera.  I love my XBox but Sony is a formidable competitor and will not yeild easily. 

But what about the growth?
The growth of computers and peripheral equipment over the past 4 quarters has been 6%, per the US Bureau of Economic Analysis.  The growth of Microsoft's Client was 7.1%, and Information Worker (Office) was 5.5%.  So the vast majority of profit associated with Microsoft is tied to the rate of expansion of PC's.  I know the US numbers don't account for China and India, but the wealthy western nations are saturated with PC's and are basically at a replacement run rate.

This name is a safe place to park your money if you have to have software exposure and believe the rest of the sector is about to disintegrate.  Otherwise, owning this stock is a head-scratcher.

Shakeup at Sun

So Jonathan Schwartz gets the top job at Sun, eh? 

The spin-machines will point out that Jon "open sourced Solaris and bowed to Linux" among other savvy strategic moves.  But I recall sitting in JavaOne with Jon on stage literally arguing with the attendees during his Q&A session.  Pretty much everyone in the audience was telling him they were interested in open source and Jon was retorting with non sequiturs like, "Java is #1 on mobile phones!"  He was obstinate, combative, and off point.   It was surreal, and this guy is running Sun? 

Granted, this was about three years ago and he may have figured things out a bit, but I doubt it. 

Much Ado

By now you no doubt have heard the open musings of Larry Ellison about creating his own Linux distribution.  Or perhaps buying one of the distributions out there- Novell or Red Hat.  What to make of all this?

Oracle has enjoyed a strong relationship with Red Hat, and was a significant early endorser of Linux at a time when there was even more uncertainty about open source than there is today.  All key Oracle products run on both SuSE and Red Hat Enterprise License.  Oracle saw the rise of open source as a strategic threat to Microsoft and has done much to feed that fire. 

Of course, lacking the OS layer in their computing stack must gall Larry to have made public comments about adding an Oracle Linux distribution.  But let's look at it from a customer's perspective.  What value is there to Oracle providing an offering at every tier of the stack? 

A customer might find value in the "one throat to choke" model, where a single vendor is responsible for the OS, RDBMS, Middleware, and Applications.  On the other had, a customer might resist the lack of competition between vendors and (rightly) fear paying higher prices across the board over time. 

On the plus side, a soup-to-nuts Oracle stack might be optimized for certain technologies such as massive parallelism (grids) or RAC.  A best-of-breed approach would not be impossible, but it wouldn't be easy either.  On the other hand, the open source community could create the same capabilities if there was incentive to do so. 

But what would compel a customer to move away from Novell or Red Hat to an Oracle distribution?  If you put yourself in the shoes of a CIO your two top goals are: 1) uptime and 2) meeting your SLA's.  Everything else comes after that.  There isn't much upside in swapping out your reliable Red Hat OS in favor of Oracle's, just because Larry said you should. 

Oracle would have to financially motivate customers to do that by bundling the distribution in enterprise deals with nice discounts and ample support.  This might work, but obviously would be costly. 

Another challenge to an Oracle Linux distribution would be the ever-increasing number of distributions.  Eventually 3rd party software and hardware companies will cry uncle.  The entire certification process is a factorial problem when you consider all the players at all the tiers of the stack.  Adding yet another OS to the picture can not be attractive from a 3rd party certification standpoint. 

But Oracle's muscle and heft would make their distribution viable.  It might not set the world on fire, but it would not be insignificant either.  The contributions would benefit the open source community as well. 

Next: Would Oracle buy Novell?  Red Hat?

Bullish on Infrastructure

IBM reported earnings yesterday, and despite my disregard for their accountability when it comes to their software numbers it is hard to believe they would be so blatantly disingenuous to pad their middleware numbers so much. 

In short: year over year growth for the WebSphere brand was 26% (30% in constant currency).  I take this as a bullish sign for BEA Systems, Tibco, and webMethods. 

But wait, you say.  IBM competes with those guys.  Those growth numbers come at the expense of the companies you just mentioned.  To which I say, that is somewhat true but counterbalanced by anecdotal evidence suggesting that there is a broad-based rising tide that will lift all boats. 

There appears to be a loosening of the purse strings in the CIO office this year, the likes of which have not been seen in five years.  And infrastructure seems to be the area with the most pent-up demand.  During the dire years, approved projects typically had rapid and demonstrable value to their financial decision makers.  It is harder to show "rapid and demonstrable" value with infrastructure than, say, applications or eBusiness solutions.  While all software withered from 2001-2004, infrastructure took it hardest on the chin.

But just as a house will collapse without strong architecture, plumbing, and wiring, so too will a given company's IT system without investment in infrastructure.  If you accept my argument that infrastructure spending has lagged overall software spending, then you can see how there is pent up demand for solutions provided by the aforementioned names.  In sum: business demands have not slackened but the spend on the enabling solutions has.  The correction, it seems, is upon us. 

The magnitude of the correction will not remind anyone of 1999, but my guess is that 2005 will be a healthy year for software.

Diversion

So there I was, perusing the various and sundry websites that provide glimpses of insight to the software world when this headline appeared:

"Scientists Simulate Black Hole Collision"

Yes, checking this link out will provide prose that includes run of the mill phrases such as "rend the fabric of space and time," and "space-torturing gravity waves."  And unlike software marketing, they actually mean this literally. 

EIght SaaS Myths Debunked

A bullish SaaS analyst walks through eight myths about On Demand software and offers some nice insights here. 

The End of the M&A Road in Sight for Oracle? Go long?

The Financial Times paper edition today has a story where Larry Ellison is quoted as musing aloud that it makes sense for Oracle to own the entire stack, including a Linux distribution.  Novell and Red Hat are mentioned as takeover targets.  Story here.  And here. 

I keep wanting to buy Oracle at these levels because I hope the M&A binge is nearing the end and I theorize that the discount applied to their stock due to the integration risk will melt away.  That alone should buoy the stock, and God help all shorts if they put together a complete quarter for the first time in ages (succeed with Tech AND Apps in NA AND EMEA)! 

Let's say they do buy Novell or Red Hat.  That leaves BEA Systems as the last obvious mega-purchase, no?  But Alfred Chuang, CEO of BEA, simply refuses to sell to Larry which means the only avenue would be hostile which would most likely be a non-starter in Redwood Shores. 

So longing Oracle here might be the cheapest it will be for a while due to the M&A discount and the looming bogeys of a Linux deal or a BEA deal.  I think the latter is unlikely for a while, and the former a real possibility.  It makes sense to me to take a small position in Oracle at these levels, then add to it after they take out Novell.

A SaaS bear growls

Lee Gomes of the WSJ Online writes a critical piece about SaaS.  (Subscription required)  His central theme is that

"Software as a service" is one of those trends that rolls through the technology world every few years, promising an end to everything that people don't like about business as usual in the multibillion-dollar world of corporate technology. Typically, these new approaches deliver on some, but not all, of the promises, often while creating new problems of their own.

He cites the well discussed issue that SaaS applications are not fit for the enterprise.  He uses Salesforce.com as his exemplar and cites a survey from AMR that indicates 36% of SaaS customers found the experience fell below expectations.  I would very much like to hear more about WHY this happened.  Was it a simple expectation mismatch?  After all, SaaS applications market themselves as brain-dead-simple.  It seems that 36% did not find their implementation brain-dead-simple.  Put another way, 64% did find it that simple.  Maybe this is good news for SaaS!

Perhaps the 36% were all Fortune 500 accounts, and this was localized to large enterprises.  But we aren't given that insight.  Mr. Gomes cites the number of SI partners on Salesforce.com's web page as evidence that these implementations are quite hairy and require professional services to deliver the return.  My take is that a complex enterprise will certainly want things a certain way, and there will always be a home for some services.  What would be telling is the average SI bill for an enterprise SaaS install compared with the traditional model?  Without that data, it's just not possible to make a bear call on SaaS based on these premises. 

The fact is, SI's such as Accenture and BearingPoint were huge off-balance sheet sales forces for Siebel (or SAP or anyone else for that matter).  Salesforce.com and other SaaS providers will need to figure out a way to motivate SI's not to sell competitive offerings, despite having a value proposition that directly cuts revenue out of an SI's pocket.  It is a conundrum that faces all On Demand software providers: what to do with the powerful players you basically deprive of revenues? 

The stakes for this argument are high, as Salesforce.com is a growth stock and depends on powering into and through the enterprise sale.  If they cannot, they will not sustain their valuation and will crash.  Stay tuned.

Chercher de l'argent!

Marc Fleury was not always so enamored of Red Hat, eh?

(Hat tip: Mongo)

JBoss finds financial exit

Red Hat acquires JBoss in a $350M in a cash and stock transaction, which could have an additional $70M included if JBoss reaches certain performance goals.  That's about $140M in cash and $210 in stock, which is about a 17.5x multiple on revenues (estimated) of $20M. 

The bulls are running with this as RHAT is up 8.9% as of 12:30 EST to $29.96.  Speculators are hoping NOVL will do something similar and the stock is up 1.8% to $7.60. 

The deal makes a lot of sense for Red Hat, but is not without some challenges.  Red Hat's application server was announced in 2004 and failed to make any sort of discernable impact while JBoss reached a critical mass of distribution and downloads.  Red Hat enhances their value and defensibility by building an open source stack, and JBoss is a nice arrow in their quiver - not to mention it deprives IBM, Novell, HP, or anyone else from that same arrow. 

The main challenge is how will these two very different open source companies gel?  Red Hat is a "purer" open source model in that they have very low R&D and S&M costs because they leverage the existing Linux developer model.  They do not spend a ton of money writing their flavor of linux, rather they certify and support a version that the community created.  Sure, they do write some code but the bulk is off balance-sheet R&D.  By the same token, they have a low cost distribution model that does not rely on Red Hat employed sales people. 

JBoss, in contrast, looks a lot more like a traditional software company in that they do most of the writing of their product (about 80% of the code is written by JBoss JEMS employees according to my notes from their recent roadshow) which has two implications.  The first is that they have some developer resources they can absolutely not afford to lose.  I am sure Red Hat locked these men and women up, but you can't account for productivity and attitude.  The second implication is the higher R&D cost associated with JBoss.  This cost will also be matched by the higher S&M cost due to the enterprise salesforce JBoss has.  Between the two, RHAT will need to make a call- do they try to cut those costs or do they leave JBoss entirely alone?  And beneath that is the fundamental issue between two companies with two models and two cultures that are not as similar as some might say.  If they opt to cut costs they will have to open up the development of the JBoss code to the entire universe, and change the entire sales distribution model- something that is fundamentally opposite of how JBoss has found success to date. 

But before I make a bear call on this, let me say that I think that even assuming some bumps in the road of integration, Red Hat will do fine.  What I understand less is why JBoss went this way.

One commenter to this site, Mongo, said there was no way JBoss would go IPO and he get's the "I told ya so" high ground today.  I would love him/her to post more on that.  I saw a chance to tap the excess demand for pure-pay open source companies in the capital markets.  Some have pointed out to me that there is so much capital available today that RHAT would not have been negatively impacted and JBoss the publicly traded company would have no assurance of a premium valuation just for being a pure play.  But if they are eating BEA's lunch like they claim they are, why not go public and build out your stack with the proceeds?  At Red Hat they get a reduced role in determining their future in exchange for another slice of the stack as well as stability and cash. 

To me, this indicates that JBoss did not have the confidence or faith in their ability to go horizontal and vertical.  I suppose this is the safe move for them, but I always associated JBoss with derring-do and a brash "us vs. the world" mentality.  Perhaps that was all marketing!

JBoss looks for financial exit

After the excitement around Oracle buying JBoss was fading, I happened to attend a conference put on by JBoss that focused on their SOA play.  And as was sure to happen, during the Q&A someone asked about the rumors that JBoss was to be eaten up by Hewlett Packard or Oracle or someone else. 
The answer was pretty smooth.  The senior JBoss person there was Shaun Connolly, the VP of Product Management, and he had three points to make:

1. JBoss accepted VC money
2. JBoss was obliged to deliver a return on that money
3. JBoss therefore always entertained serious inquiries about being acquired, but they were preparing for an IPO.

This article today basically says the same thing.  Which leads me to consider JBoss the publicly-traded company.  My gut reaction is that the market would jump on this with both feet and go bananas.  As mentioned here (and in a million other places), Open Source is arguably THE disruptive trend in software today.  However there are just two plays for institutional money: Red Hat and Novell.  Just taking a gander at Red Hat's P:E (69) compared to software in general (32) or the S&P (22). 

Novell, on the other hand, is a mess.  As in, Messman, their CEO.  I have the dubious experience of working for that guy ever-so-briefly at Cambridge Technology Partners.  He did not impress there, nor do I hear good vibes for him coming out of Novell.  One of my many goals is to unravel the mystery how Novell can have a nice maintenance cash stream coming in coupled with ownership of some nice open source properties and have been met by a deafening silence in the market? 

JBoss has the ingredients to potentially be successful. 

• Pent up demand for pure-play open source stocks
• Profitability
• Comfortable installed base

On the downside, once they file, they will hop on the 90 day treadmill which will be interesting to watch.  Thus far they've been able to throw slings and arrows at the likes of BEA and IBM from the safety of not having to disclose anything about their own performance.  It will be instructive to see how Fleury handles his first miss, or if some of their past behavior comes under harsher scrutiny. 

Top 10 IT Projects

This survey posts the top 10 IT projects by projected spend.  Interesting.  I wonder, is this what salespeople are seeing?

The Quickening?

Obviously I'm an open source bull, and I think of Dan Patrick's line when I consider the current state of Open Source in the software market: "You can't stop it, you can only hope to contain it."  Surprisingly cheesy and accurate at the same time.

There is an article in eWeek today suggesting that Microsoft is about to go to the mattresses with Open Source by attacking the IP via their hoard of patents.  I think this would be a good thing ultimately for the Open Source community because it would clear the air of this implied threat that Microsoft makes from time to time.

It is hard to guess how much headwind this threat creates for Open Source. My guess is: not much. The heart of the community are famous for taking on commercial vendors and probably will not make the distinction between a code war and a legal battle.  In any case, I predict a victory for open source in a legal battle because I ultimately think the community will take the offending code and re-write it under an open license.  As a practical matter, who is Microsoft going to sue? Red Hat? Novell?  VA Linux (ha!). The community is just to decentralized and ownership is diffused literally across the entire globe.  Even if Microsoft destroyed all US-based open source companies, they'd re-emerge elsewhere around the globe, outside the enforcing reach of the USDOJ.

Even I notice that I am not taking a stand on the legal merit of such a move by Microsoft.  Groklaw can do that.  To me, the practical realities seem beyond surmountable. 

If my prediction is right, then this battle will be a temporary containment of open source, and will pave the way for an even more decentralized and difficult to stop movement.

In a future post, I will contemplate the question: "So who makes money off all this anyway?"

Java's evolution

Full disclaimer time: I'm not a programmer.  I have never laid down a line of code in my life.  So I'm going to expose myself to mistakes of ignorance with this next post, but I do so to test my hypothesis in the wild.  I'm walking through the bazaar with a bolt of cloth, trying to see if there are buyers!

In the 1990S, Java triumphed over some alternative languages such as Vignette's TCL and ATG's Nucleus  Both TCL and Nucleus were proprietary but shared the virtues of Ruby on Rails in that they were far lighter/faster/more flexible than Java.  So why did Java triumph?  And why is Ruby eating it's lunch, according to this story. 

Here's my take: In the pre-Java world, the software revenue model was dominated by firms writing proprietary code.  This code delivered value to customers, and the software vendors reaped excess profits.  In pure economic theory, anytime there is excess profits other vendors enter the market offering the same value until there are no excess profits anymore.  The only way to prevent this, if you are the original vendor, is to erect barriers to entry.  Proprietary code was a significant barrier.  Now you can ensure that a rival vendor's products will not work well (heck, at all!) with your products and your first mover advantage should mean you stay ahead in the arms race for new value-adding features.

Remember, in the pre-Java world, the software was tied to the hardware via the proprietary code.  So not only did you get locked in from the software side, but you also got locked into a vendor from the hardware side. 

<sidebar> If I ever get a law named after me, I hope it's this: "The market abhors lock-in."  Moore has his law.  This is mine.  Call it "Reed's Law."  The corollary is this: "And the market will ALWAYS find a way around being locked in, in the long run."  </sidebar>

Enter Java, ballyhooed as "Write Once, Run Anywhere!"  This may sound quaint today, but this concept was quite exciting to CIO's around the globe.  And Java established a control organization and process called JCP (Java Community Process) that let any vendor propose a direction/feature for Java which was evaluated and either accepted or denied by an executive board comprised of some of the heavyweights - IBM, Sun were but two.  This gained a tremendous following with an almost religious fervor to it.  For the first time (I believe) the IT landscape had a programming language that was independent of the hardware below it, did not have one vendor dictating future direction, adhered to open standards, had a critical mass of developer support, and the marketing engines of Sun and IBM cranking out the PR. 

While Java is the catch-all term for this open language, the version of it aimed at enterprise computing requirements is called Java 2 Enterprise Edition, or J2EE for short. 

When considered this way, it's actually impressive that Microsoft got as much of the market with Dot NET.  The debate between J2EE and .NET has jumped the shark with the consensus being that J2EE is better in multiplatform environments, and .NET is superior in Microsoft-centric IT shops. 

As mentioned elsewhere on the blog, J2EE was designed for industrial-grade computing needs and as such is not as useful for lower-level enterprise computing needs.  This is the area where TCL and Nucleus should have thrived, but didn't.  I would argue that their proprietary nature got them killed off.  And today there are a bunch of new, dynamic languages that are truly open source.  Python, PHP, and Ruby on Rails to name a few.  These are entering the vacuum created by J2EE and are like rain in the desert.  They are easier to use, more flexible, and can handle a large chunk of application demand. 

Ultimately, the landscape will stabilize, and J2EE will not go the way of Cobol (as some have mentioned).  The CIO now has a more diverse pallet of programing language options to address demands from the business side.  For massive transactions with no tolerance for failure, J2EE is a good choice.  For banging out a bunch of pages for a social networking site, PHP is a more popular choice. 

The evolution of IT continues, and the winners are the people who have demanded interoperability, speed, ease of use, and performance.  These aren't new demands, and it's a rumination for another time why on earth it has taken this long to even get this far?  The software landscape is about to be leveled by open source and on demand... the future will be quite different than the past. 

IDC: Vista Delay Won't Hurt PC Shipments

eWeek cites an IDC report that contends that PC shipments in 2006 will not be significantly impacted by the delay of Vista.  Money quote:

"The decision will have some impact on PC shipments, but it will be limited, as businesses aren't likely to change their plans-most will evaluate the OS for at least a year before rolling it out-following the change. Consumer PCs will feel the brunt of the blow. But even that could be smaller than originally thought."

Does this mean that the world is saturated with personal computers?  Or that the marginal features in desktop operating systems do not add much value anymore?  My guess: both. 

As long as Microsoft is still principally a desktop OS/productivity suite play, it is a value trap.  The shining star is the server side products, and the promising divisions are Dynamics and Xbox.  Anyone out there with an opinion of their MSN business?  My gut says it could be a cash waterfall for them but they have consistently been a day late and a dollar short compared to Yahoo and Google. 

Tibco's Q1 Earnings

Tibco reported earnings after the market close yesterday for their first fiscal quarter of 2006.  The numbers beat guidance and analyst EPS estimates, yet traded off on an up market day.  Tibco posted EPS of $0.06 versus expectations of $0.05, and revenue of $114.6M versus consensus of $110 and a guided range of $109.2 - $112.5M.  Operating margins were 14.7%, cash flow from operations was $30.8M.  About the only negative point was that deferred revenue declined approximately $1.03M, versus increases of $12M in Q4 and $10M in Q105. 

Days like this will make TIBX shareholders pull their hair out.  Credit Suisse's Jason Maynard observes:

"For the third quarter in a row, Tibco beat estimates without closing any mega deals. It appears business has stabilized to the point where the company can continue meeting estimates without the help of large deals... While we had no expectation for Q1 to be a “beat and raise” type of quarter, we were generally pleased to see solid operational results across the board...  As we progress through 2006, we continue to believe there are a number of potential catalysts that could accelerate growth, including improving sales execution, robust demand in key verticals (energy, financial, telco), and the potential for some large deal contribution.

Tibco's backbone historically has been their superlative messaging product, Rendezvous.  This product has commanded a premium in vertical markets that spend more on technically proficient solutions (Financial Services, Energy, Telco).  When I began covering Tibco in 2003, they carried a stigma that they relied too much on $5M or even $10M deals, and this created volatility in the name.  Hence Jason's point about the ability to beat expectations without resorting to an eight digit telco deal in Europe, (for example).

Even though the mega deals have been absent for the past three quarters, it is noteworthy to point out the ASP for deals over $100K was $725K for the quarter.  This follows the past three quarters trend of $450K, $500K, and $600K respectively.  CEO Vivek Ranadive did remark that Tibco was able to maintain pricing power this quarter.  For the Tibco bulls, this could be a sign of strength - the company is improving their execution on large deals.  The bears will have to acknowledge the success and look to see if this data point is a fluke. 

All in all, Tibco is a stock to play if you are a believer in SOA.  They have the messaging and the business process orchestration.  They do have some wood to chop on the web service IDE (for lack of a better term), and I have to dig a bit on how compliant they are with the key web services standards- I assume they're there with the key ones.  But what they have is enough to more than outstrip the rising tide that SOA claims to be. 

More on SOA later.  What is it?  Why now?  Is it the real deal?

Microsoft's SaaSy CRM

At their annual ERP customer event, Microsoft announced a hosted version of their CRM product making them the latest entrant to the SaaS market.  William Gates had the requisite ringing endorsement:

"Software as a service [SAAS] is a very important trend," said Microsoft Chairman Bill Gates during his Monday morning keynote at the Convergence conference here. "We believe in it a lot."

Microsoft gets credit for opening up this model to the market.  There are a few wrinkles though.  First and foremost, Microsoft is a partner-centric company.  This point should not be understated and deserves a little more time: Microsoft is a quantum leap beyond any other major software vendor when it comes to partnering.  Not BEA, not IBM, not Oracle can even touch them.  The culture and model are partner-first.  So how can a partner strategy lend itself to the channel? 

It appears that Microsoft is letting the partners do the hosting.  If this is accurate, this creates multiple instances of the server code across the globe.  Synchronizing them all will not be easy, and I will go out on a limb and say that the participatory development model of continuous updates will not be a major reality.  So you get many static versions of Dynamics CRM Live, one per service partner.  Some of the benefits conferred by hosting disappear, principally the ability to update or patch the server software once and let all users benefit from that patch immediately.  This isn't so bad to the end user who only wants to access the product through the browser.  But it is not as agile as a single-source multi-tenant model such as Salesforce.com. 

More on the practical implications of this later.

Brittanica and Wikipedia

Read this story to get up to speed on a fascinating milestone on the way to "Web 2.0."  In short, a wildly biased report originally claimed that Wikipedia and Brittanica were about the same from a fact-quality point of view.  However, under further review, the article was closer to fiction than not. 

Buried in the facts of the case is the point that Wikipedia had fixed their errors (no word on Brittanica).  I am struck time and again at how quickly the collective will and intelligence of Open Source can clean up their messes. 

Ultimately this should serve as a guidepost on when to use the slower but more reliable offer versus the quick and "good enough" alternative.  They aren't mutually exclusive, but the focus will be on the bazaar in the near term because IT has been all Cathedral for ages.  The additional granularity of service level in software is a good thing for CIO's who may be able to match requirements with the solution that has the right price/feature balance. 

Metering

John Udell is onto at least one of the flies in the ointment of distributed, composite applications.  Money quote:

"When we talk about a grid of Web services, we like to compare it to the power grid, but the analogy is deeply flawed in at least one way. My electric bill isn’t itemized."

While I think that composite applications are an almost certainty, several kinks need to be worked out.  The Off-The-Top-Of-My-Head Top 3 are:

1. Uptime - this needs to be many nines
2. SLA's - how do we ensure these from across many different providers?
3. Metering - how do we charge for use?

And all the while, composite applications must be easy to assemble and use.