When I started Ponderings, I hoped to re-connect with old colleagues and more importantly expand my collaborative keiretsu in new and interesting ways. Without question this experience has far exceeded my expectations on both fronts.
The greatest byproduct of this social experiment has been the formation of the Enterprise Irregulars. What started as a loosely coupled group of bloggers who had nothing more in common than knowing Jeff Nolan, has evolved over the last few months into something much, much more.
Originally coined the Irregulars by Erik Keller following our SAP Sapphire attendance, the group is far more expansive than that and has come to include enterprise software executives, consultants (at big shops and of the more entrepreneurial variety), investors (yours truly included) and journalists. We keep in regular contact via an email group, Crispy News, a wiki and group conference calls. We're also looking at ways to further collaborate, including a regular series of podcasts.
Tonight the Irregulars make our first collective foray into the op-ed world in response to last week's Sandhill.com op-ed piece, Is Enterprise Software Doomed?
The piece, written by Guy Smith of Silicon Strategies Marketing, made some bold statements about the impact that open source and SaaS were having on the enterprise software industry and the incumbent providers. The piece elicited a passionate response from many of us; and the good folks at Sandhill were kind enough (via Sadagopan and Erik) to offer us a chance to counter some of Guy's claims.
Our retort, Software's Sky is Not Falling, appears tonight and encapsulates ten distinct viewpoints on Guy's initial take. As you'll see in the op-ed piece, we approach this topic in a lot of different ways; but all agree that Guy's contention that "With little effort a commodity stack can be deployed for 95% of all IT buyers" is quite debatable.
Due to space constraints, only a portion of each of our op-ed retorts made the final article on Sandhill. In the interest of full disclosure and furthering the conversation, I'm posting the entirety of my submitted response. I look forward to hearing your viewpoints. Do you agree with the Irregulars viewpoint(s) or are you squarely in Guy Smith's camp? Somewhere in between?
Sandhill recently published an Op-Ed piece by Guy Smith entitled, “Is Enterprise Software Doomed?” Pithy as that may be, I think Guy obscures some astute observations with unjustified hyperbole.
Of particular issue is his view that the IT stack can be completely commoditized. In his piece, Guy writes:
The entire IT stack can now be commoditized. With little effort a commodity stack can be deployed for 95 percent of all IT buyers. If you think this an absurd view, ponder for a moment what has already happened to the stack.
1) Au contraire, it IS an absurd view
2) I’ll happily offer my thoughts as to where I disagree with Guy
Smith offers a rudimentary outline of the “stack” and cites examples that purport to support his notion that we’re facing 95% commoditization.
Smith on Servers: Giants like e*Trade, Amazon, Yahoo, Google, and others are using x86 and x64 servers on everything except database hubs, and even those are targets for eventual replacement with massive multi-core x64 boxes. Little guys are using x86 servers universally.
My response: Servers are certainly becoming commoditized, but Smith plays bait and switch here. He’s referencing the commoditization of hardware and microprocessors when he’s supposedly making an argument for the demise of enterprise software. Hardware does not equal software.
Smith on Operating systems: For new deployments, it is either Windows or Linux, and Linux - along with Microsoft's tardiness on Vista - has caused Microsoft server sales to stagnate. Linux was born a commodity, and remains such.
My response: The numbers don’t bear this out. Microsoft SQL Server is taking share alongside Linux, both at the expense of Unix. Furthermore, I have trouble accepting that Linux OS is really a commodity in the traditional sense. At the enterprise level, Red Hat Enterprise Edition is the pervasive option and that’s really not a commodity product. Red Hat’s Linux OS business has 24% operating margins last quarter (up from 12% a year ago); you would be hard pressed to find anyone who would define that as a pure commodity. Red Hat has leveraged an open source code base and made it its own. They’ve layered high end (and profitable) services that make it appealing to enterprises. How many companies do you know that simply run their own versions of the Linux OS kernel?
Smith on Application servers: When Apache instantly dominated the Web server market, and started chipping away at the Java server market, once-mighty competitors started reeling.
My response: I’m confused here…LAMP is certainly massively popular, but since when have Java server vendors started reeling? Oracle’s J2EE app server has become a $1 billion business in just two years, and last time I checked Oracle was still printing money with 40% operating margins. Meanwhile BEA continues to chug along thanks to a refined marketing message around SOA. Ironically, JBoss, the flag bearer for enterprise class open source app servers, is the one struggling. Red Hat (which acquired JBoss recently) just guided to a $0.04 per share negative impact from JBoss in FY07.
Smith on Application development: With IT shops highly favoring Web-based application development for new projects, we see PHP, Perl and other Open Source systems replacing proprietary tool sets (what was the last 4th GL you sold?)
My response: PHP and Perl are doing quite well, but since when has enterprise software explicitly made money from its scripting languages? Meanwhile the application testing and development tools market (Mercury Interactive being the poster child) continues to grow profitably and is arguably the hottest area for M&A consolidation.
Smith on Databases: Oracle and IBM have retreated to the high ground, and gone as far as offering free versions of their DBMSs to low-end users, in an attempt to forestall adoption of commodity database systems like MySQL, PosgreSQL, and others.
My response: I’m a fan of Ingres, and believe they have an interesting message to tell the market. That said, how can Oracle be accused of “retreating to the high end” at the same time offering free product to the low end? The model of providing a low-cost/free version of core functionality to seed the market and sate the low end is nothing new; if anything that’s good business. Again, look at Oracle’s margins or IBM’s software margins and tell me they’re being commoditized.
Smith on Applications: The last refuge of high margin software is coming under attack in segments like CRM (SugarCRM, Compiere, OpenSourceCRM), ERP (Compiere, ERP5, OpenMFG), messaging and collaboration (SendMail, PostFix, Open-Xchange, Astrisk), verticals (MedSphere, OpenEMed, OpenClovis) and horizontals (Open Office, Plone, Zope).
My response: Maybe Guy Smith is talking to different corporate customers than I am, but every time I get a peek into what a CIO is doing, or looking to do I frankly don’t hear these vendors mentioned. One has to be intrigued by what Sugar and Compiere and others are doing; but it’s a long, long way from making a dent into the market share and installed base at the enterprise level. Even in the mid-market (which I define as $500mm revenues +), these solutions are still very much in a “show me” state.
Guy then goes on to discuss his view on the trifurcation of the software industry into three buckets: Traditional, Open Source, and SaaS (Software as a Service). Obviously he’s oversimplifying here (and I’m sure he would acknowledge that), but again I find some of his conclusions contrary to the evidence before us.
Smith says: The not-so-big secret is that 95 percent of all IT needs can be served by commodity solutions. This is where Open Source firms make their killing.
My response: Again, the 95% number is way off base. And as to Open Source firms making a killing, can we define that a bit more? Is he speaking of building a sustainably profitable business model? Because I don’t know of too many OS companies that carry that banner. Red Hat is one, but again, they’re very much a hybrid model (using open source as the base, but extending it in a proprietary way). How many OS companies are profitable or close to being profitable?
Smith says: SaaS is the bastard child of the traditional proprietary software vendor and the Open Source marketing paradigm. SaaS offering a typically proprietary, but are marketed using many of the same processes as Open Source (PR, buzz, online sales cycle process management, and so on).
My response: SaaS is hardly the bastard child of open source; it’s evolved on its own. Admittedly they have followed parallel paths (both creating a market out of the inefficiencies of traditional enterprise software) but SaaS is not a byproduct of the OS movement. SaaS and Open Source are more like brothers; both evolving from traditional software.
At the end of the day, I think Guy Smith is right to point out that traditional enterprise software has challenges; important ones. What open source and SaaS have done, in my opinion, is make customers aware of their options. They are now starting to question the ROI of 17%-22% annual maintenance, and the vendor lock-in that is created by purchasing perpetual licenses and then spending tons of money on customizations that have to be re-written every time you upgrade to a new version. But we’ve been down this road before. It wasn’t long ago that the ERP vendors were left for dead as the new best-of-breed vendors (CRM, supply chain, etc…) were going to steal away market share permanently. We know how that ended. There will be revolutionary and vitally important companies that emerge from the open source and SaaS models to be sure. But they will only do so if THEY figure out how to look MORE like the existing vendors, as much as the traditional vendors need to learn a few things from them.