Randall Stross tried to write an interested column in the Sunday New York Times about a real problem — the declining growth of IT spending — but missed an important point. Two, really. The first is that he makes the common, business page mistake of lumping ALL spending related to technology and business into a single bucket called IT spending — that is hardware, software, services, PCs, mainframes, blade servers, and whatever else there may be in the IT kitchen sink. The result is an overbroad analytical brush that obscures some important truths about IT spending and growth. The second is that, in the process of making his shaky argument, Stross takes CRM-pioneer Tom Siebel’s suddenly humble word for it when Tom claims that the glory days of IT spending are past.
Beyond proving that two wrongs don’t make a right, Stross’s first sin is one he shares with the entire financial community, the New York Times among them. There is nothing sillier to see than the sector charts that papers like the NYT put out, which lump semiconductor companies, consumer software companies, Google and IBM in the same sector. This is, unfortunately, the IT market that Stross (and Siebel) seem to be claiming is in fatal decline: a sector so broad and complex that any statement about its performance as a whole would be, well, largely absurd.
Take just the software side of this IT market that Stross and Siebel think is in decline. The Thomson Reuter people who build the charts for the NYT define the software sector as “companies engaged in developing and marketing system and application software. The industry includes developers of operating systems, word processors, spreadsheet applications, CAD and database engines. The Software industry excludes applications customized for specific tasks requiring continuous support from developers.
Got that? CAD, wordprocessors and spreadsheets, and, by the way, enterprise software. TR’s list of companies that fit this bill runs into the 100s, too many to list here but enough to show that their over-broad definitions of software alone make software sector analysis a non-starter. And when you include the rest of IT into this analysis — in addition to semi-conductors, you can add office equipment and telecommunications equipment as well — you now have a potpourri of “stuff” that has, for the most part, has little in common other than the requirement for an electrical current in order to useful.
This makes for some rather ill-informed analysis about what is actually happening in the real world about what products are selling and who is buying. Clearly, anyone who says “IT spending” is going one way or the other, based on this over-large bucket o’ technology, is making a statement that lacks any useful precision. If we use the Thomson Reuters hierarchy — and apply a little rationality to the criteria for analysis — you need to drill three layers deep in order to start to say something meaningul. You can’t really talk about IT as a single class of company, nor can you, using the TR hierarchy, talk about software either. The analysis has to be much more fine-grained than these broad sectors define: enterprise software, CAD software, PLM software: those are categories that bear the kind of analysis that Stross and Siebel are trying to make. Talking about IT without specifying enterprise software or CAD software or semiconductor manufacturing or spreadsheet software makes any discussion of IT a little silly.
So, one point against Stross, who generally gets kudos from me for his insights. Sorry.
Stross loses point #2 when he takes so much of what Tom Siebel tells him about Siebel’s history in the software industry at face value. First, he characterizes Siebel as “self-deprecating” and “modest”. Anyone who knows Siebel or ever interacted with him or his eponymous company would question whether Stross was even talking to the real Tom Siebel, or some well-medicated stand-in. The real Siebel was, well, anything but self-deprecating and modest, to put it very mildly.
More importantly, Stross allows Siebel to claim that, because “the promise of the post-industrial society has been realized”, IT spending only grew by a rate of 3 percent since 2000. The statement is as ludicrous as the “end of history” assertions that followed the collapse of the Soviet Union, something that we all know in the 9/11 era was a pretty dumb idea. Anyone, even Tom Siebel and Randall Stross could do it, could walk into any company in the world and find what I call the proverbial innovation opportunity: a bunch of men and women working out a business problem using white boards, spreadsheets, speakerphones, and sneakernet. As someone who works extensively with start-ups and early stage companies, these “white space” opportunities are the crucible for a seemingly endless incentive for innovation. The promise of post-industrial society is still a long way from being realized. That fatal error — the end of innovation — was controversial enough to give Nick Carr a career he otherwise might not have had, but that doesn’t make it true, neither today nor for the foreseeable future.
To Stross’ credit, he quotes several other pundits, including Timothy Bresnahan of Stanford, who makes pretty much the same points about innovation that I made above. And the conclusion of the column is that Siebel’s premise may be wrong, though Stross continues to base his analysis on the same falacious assumption that IT spending is a granular-enough data point for serious trend analysis.
In the end, perhaps the biggest sin was listening to Tom Siebel in the first place. By the time he bailed out of his eponymous company by selling it to arch-rival Oracle, he had clearly lost his visionary cred: Siebel missed the SaaS market, losing it to Salesforce.com, and he missed multiple opportunities to broaden his strategic footprint and present a more comprehensive market solution than just CRM software. And, perhaps even more important, this isnt’ the first time Siebel has opined that tech investing is a dead letter: he did that in June in the San Jose Mercury News. My guess is that, in addition to promoting clean tech as a replacement for high tech, Siebel is actively shorting the tech market. That would be more consistent with the Tom Siebel I crossed swords with in the day. The newly self-deprecating and modest Tom Siebel — and his arguments against the glory days of IT — just don’t make sense.