I'm constantly astounded by the amount of attention given to Microsoft's monopoly threat, which, contrary to nearly all popular wisdom, is completely inconsequential.

Let me make it absolutely clear to anyone running a business or an IT department: Microsoft's lock on the operating systems market does not matter to the future of your business in the least. Never has. Never will.

Let me tell you about the monopoly threat that does matter to the future of your business: SAP.

Bill Gates may be much more wealthy than Hasso Plattner (who?), the hyperkinetic German who started SAP in 1972 out of frustration with his American bosses at IBM who wouldn't listen to his plans to develop systems that would automate the major activities of a business in software. But Plattner, who is now retired as Gates is threatening to do, has had an infinitely larger impact upon the ways that IT and businesses are run than Gates ever has.

That's because PCs and operating systems have nothing to do with how a company operates. SAP's software, meanwhile, has become the basis upon which many companies go to market and compete. SAP is their business, written in software. And, as the vendor broadens its reach and captures ever more share of what is known as the enterprise applications market, SAP's strategy has come to define its customers' strategy. Many of these companies have literally abdicated their business strategy and handed it to their software vendor.

It's not hard to do. The founders of Google could never have inspired their troops by saying they wouldn't act like SAP. (C'mon, everyone knows that "Do no evil" is a direct reference to Microsoft.) This is not a company putting out second-rate code and building market share by detonating competitors in ethically, if not legally, questionable ways. SAP may be the best example of what former Federal Reserve Chairman Alan Greenspan referred to in a speech last Friday at AMR Research's Boston conference as a "natural monopoly." It has gotten to where it is by simply outperforming its rivals in every way in a market that favors scale, quality and reliability over all else.

SAP has long dominated the enterprise application market for big companies, but today, you can safely call it a monopoly in that sphere. Let's consider the numbers, all from AMR Research. In the traditional core market of enterprise applications, ERP (Enterprise Resource Planning), SAP's 2005 market share of 42 percent was more than double the 20 percent of its nearest rival, Oracle (whose share would be a lot smaller if it hadn't acquired some of its ERP rivals in recent years). All other companies, mostly niche providers serving smaller companies or specific industries, were in the single digits.

But the real story is in other areas of enterprise applications, because these demonstrate the real market power of SAP. In supply chain software, procurement software, and customer relationship management software, SAP is now number one. In HR applications, it is merely second by one percentage point to Oracle. The only market in which it lags at all is product lifecycle management. Together, these categories comprise nearly all of the major business processes of a large company.

That simply never would have happened in a normal market.

You could argue that SAP got to be number one in ERP because it had been perfecting its software since 1972, gradually building its market position over time in the traditional way. But in these other markets, SAP didn't offer any software until very recently. Yet as soon as it entered these markets, its share began rocketing to the top. Why? Because its large customers believe it is too risky to buy software from anyone else. They withhold purchasing new stuff and dismiss vendors with more complete or more targeted solutions because they fear the problems they would have integrating software from these vendors with their existing SAP software. This makes SAP's entries into new markets the software equivalent of a Beatles reunion album--there is so much latent demand for the stuff that it is guaranteed to hit the charts at number one the day it arrives.

SAP isn't the only enterprise software vendor that causes this kind of software selection cowardice. Enterprise software represents such a large percentage of the IT spend for just about any size company that once you pick a vendor, it's hard to justify going with someone else. The immaturity of IT in general, and the unwillingness of IT vendors in particular, to make their software products work with products from another company, make the risk of going with outside vendors more than a mere cost risk (extra time and money on integration, for example) they make it a business continuity risk (the two products fail when you roll out the system and the business grinds to a halt, for example) and a personal career risk for anyone involved in making the decision to pick more than one vendor (imagine orders stuck in warehouses as some yes-man asks the CEO rhetorically, "Who made the decision to bring in this other vendor, anyway?").

But SAP has been so successful at building its customer base and expanding its market reach that for big business, it is, in effect, a monopoly, even if it doesn't look or act like a monopoly. What really matters in a monopoly market is the degree to which the vendor controls--either directly or indirectly--customer choice. And for risk-averse CIOs and CEOs in the Fortune 1000, SAP is the only reasonable choice. Could they make another choice? Absolutely. If anything, SAP is too nice to its competitors. It prefers to ignore them and build all its software itself, rather than buy or kill them. As far as I know, the nastiest thing Hasso Plattner ever did to a competitor was when he reportedly mooned Larry Ellison after Ellison cut off his boat in a yachting race.

No, all SAP needs to do is to keep working hard and offer reasonably reliable software. The spineless decision makers in the Fortune 1000, who can be relied upon to avoid risk above all else, will do the rest.

This is the kind of situation Greenspan was referring to at the conference when he was asked about the increasing consolidation in the software industry today. As we have shifted from making had goods like cars and steel to more knowledge-based goods like software, the proportion of fixed costs in the products to variable costs has gone up, he said, and that situation tends to create natural monopolies. In other words, while it costs a lot to design a car, it costs even more--proportionally--to design a huge, complex chunk of software like an enterprise software application. And while you can farm out much of the manufacturing of cars to others, thereby reducing your fixed costs, the software industry hasn't matured to the point where a highly reliable supply chain has emerged to do the heavy lifting. So that means to get into a software market, you have to invest a lot of time and money inside your own company to build the software before you make a dime from customers. That requires a lot of money and a lot of patience from investors. Long ago Microsoft and Apple erected a barrier to entry to the PC operating system market that has prevented anyone--aside from people willing to work for free--from developing anything comparable. We are in the same kind of situation in the enterprise software market. SAP is Microsoft and Oracle is Apple.

Except this time, no one is willing to admit it.

That's because to do so would mean that you have abdicated not only your IT strategy, but, to a great degree, your business strategy, to a software vendor. And that's a career threatening proposition. After all, who needs a CIO or an IT department if all you're doing is installing software from a vendor? Who needs a CEO whose business processes are mirror images of the processes at a competitor who has installed the same software from the same vendor?

I heard from many installers at the AMR conference last week. One guy talked about how his IT group had gotten the time for an SAP installation down to a matter of three months in various geographies and acquisitions around the world. It didn't matter that he was honing his skills on a ten-year-old version of the software because the costs of upgrading are so huge--tens, even hundreds of millions of dollars, or as much as it cost to install the stuff in the first place--that he keeps installing old versions of the software so that it will line up with the old software they already have.

Nor did people seem particularly alarmed that they play no role in telling SAP how to develop the software that defines their business processes. Brian Nicholas, a presenter from SAB Miller, a brewing company, kept protesting, "We're not P&G. We're not big enough to have the clout to influence what SAP is doing." The largest brewing company in the world, with one of the most recognizable brands in the world (Miller) and it doesn't have any influence over its software vendor? Scary. Nicholas says he hopes to gain more influence through the advisory board that he created that includes other big food and beverage companies.

What the installers don't seem to understand is that the more they all buy the same software from the same vendor, the less say they will have in their own business processes. Again, it's not a question of SAP being evil; it's a question of market reach and depth. As the vendor covers more and more industries with more and more categories of enterprise software, the ability to drill down into those hundreds of different industries and offer specificity becomes nearly impossible. This is why there used to be hundreds of different ERP vendors, mostly arrayed by specific industry segments. Today, those numbers are in the dozens, according to AMR Research, and even with an annual R&D budget of 1.1 billion Euros, SAP is only going to have so many hours in the day.

How did we get here? It's a matter of the immaturity of IT in general.

Talking about the 20-30 year lag in productivity gains that historically occur after new technology is introduced, Greenspan recalled touring a strikingly tall and narrow factory building in his youth as a financial auditor. The operator said it was built that way because in the old days of weak and expensive steam engines, gravity was still an important part of the manufacturing process. By the time Greenspan arrived at the plant, it was full of electric motors that more than compensated for gravity. But the investment in the building had already been made. More useful buildings--and more productive use of the electric motors--would have to wait until the old building crumbled and needed to be replaced.

The similarities to today's computing infrastructure were hard to miss. The current generation of software relies too much on gravity but is too expensive to displace. When the central concern in the buying decision is avoiding complexity rather than choosing the best or most complete solution, you have a fundamental problem. Worse, though the internet may yet turn out to be our electric motor, it's still little more than an unreliable prototype. It's not yet clear what we should be moving to, but the inflexibility, expense and slowness of the current engine are abundantly clear.

This, combined with SAP's dominance and the huge investment its customers have made in the status quo are leaving a vacuum of innovation into which no self-respecting venture capitalist is willing to step. "Venture capital companies aren't investing in enterprise applications anymore; they're investing in Web 2.0 companies," said Jim Shepard, AMR's enterprise software analyst at the conference last week.

Worse, the enterprise application vendors don't have a vested interest in breakthrough innovation, according to Shepard. "It's not in their business models to innovate," he said. "They are using their money to respond to competitive threats and move into new markets but what it doesn't get used for is to write innovative niche applications." According to Bruce Richardson, another AMR analyst, it takes the big enterprise software vendors a minimum of 18 months to respond to requests for new functionality and five years before the new features are stable enough to be used in mission-critical situations.

So you can't change a business process for at least 18 months and probably not on a companywide basis for about five years. This is the real impact of monopoly in this market. Not only does innovation slow at the vendor, but it affects the ability of the customer to innovate. Microsoft Windows never had that kind of impact on business strategy.

It's unlikely that companies will wake up one day and decide that over-reliance upon a single vendor is a bad thing. "Microsoft's domination caused everyone to learn Windows," said Greenspan. "If someone came along with something new I would probably say I don't have time to learn it. Companies that get in first and define a template are difficult to dislodge."

The AMR doomsayers did say that there's a way out of this situation, but it seems that it will come only when SAP has more market reach, not less. The company's promise to rebuild its software into service-oriented chunks that will be connected together with the company's own integration middleware (yet another area of enterprise software where SAP's entry rocked the existing market) means that smaller niche vendors could have the opportunity to sell software that could be integrated less painfully into that "ecosystem." Yeah, and the Netscape browser fit well into the Windows "ecosystem."

Meanwhile, any hope that Oracle will offer a realistic challenge to SAP seems farfetched. To succeed, Oracle would have to do something that even SAP is having trouble with: convince its customers to upgrade the software. The challenge is doubly difficult for Oracle because many of its customers come from acquisitions and never used Oracle software to begin with. They all seem bent on forcing Oracle to continue to support the old software it acquired. The tall, skinny factories don't have enough cracks in the foundation yet.

Maybe it doesn't matter that your business processes will someday all be dictated by a single software vendor. We humans move slowly anyway and we're just not that smart, according to Greenspan. He noted that even in the boom years of the industrial revolution productivity gains in this country never got above three percent a year. "Human I.Q. can be about 200 and to get beyond that we'd need pharmaceuticals," he said. "Aristotle had to have the same level of intelligence as we do today. If anything, it's been downhill since then."

You do not have flash or javascript support.

Average (2 votes)

Reply