Gloomy prospects

In recent posts, I've discussed contraction in the enterprise software market, the withering of the integration business and the commoditization of IT. There's a clear theme here: I foresee difficult times ahead for the IT sector, caused by the impact of standardization and commoditization of what were previously proprietary and highly specialized technologies and skills.

Recent articles on the site have supplied evidence of these effects beginning to take hold in the market, for example:

• In Integration that defies EAI convention, we presented the use of a standards-based messaging infrastructure to "slash the cost of application integration projects by two thirds or more" by cutting out the need for an EAI platform.


• Last week, in Integration: Free at last?, we presented the EAI vendors' response to this threat. "Integration is not free with web services technologies," webMethods' Andy Astor told us. "You still require tools to do data transformation [and] routing, and you need to have a publish-and-subscribe messaging infrastructure in place." But those capabilities are becoming standardized, too, and we could find little evidence that customers will rely on EAI vendors to provide them in the future.


• An earlier article, SIs adjust to living without integration, looked at how web services affects customer relationships with systems integrators. "Instead of a $1m project to bid on, [the SI] is bidding on ten to fifteen $60k projects," IDC analyst Sophie Mayo told us. Deal sizes are down.

An article yesterday at internetnews.com, Web Services Boon for Customers, Not Investors, reinforces this message. It quotes a new report from Deutsche Bank Securities analyst William Zinsmeister: "We see few ways for investors to profit off this pending architecture shift ... We think web services simply increase the pace of software services commoditization."

In a macroeconomic sense, what's happening in IT today is a classic case of an extended boom leaving behind a dual overhang of innovation and over-investment. A large part of the innovation surviving from the boom years consists of royalty-free open source software and standardized interface technologies, while the over-investment has left enterprises with hugely underutilized IT infrastructures. Those enterprises now find they can use the royalty-free tools and interfaces to get more productive use out their existing infrastructure, for little if any marginal cost.

For the past decade, the conventional wisdom in IT circles has always been that enabling customers to save money stimulates further growth in the industry, because it creates extra profit that can then be reinvested in additional technology. Thus there's a virtuous cycle of productivity feeding investment, and even though prices are falling, customers always end up buying more.

But what if, instead of retaining their profit, customers use a lower cost base to cut their prices? In that scenario, the proceeds no longer flow back to IT vendors. Instead, the savings flow out into the wider economy, and they only come back to the IT industry when new investment starts to pick up again.

I raise this gloomy prospect not in order to spread despondency, but simply in a spirit of realism. Recognizing what the future might hold is the first step towards dealing with it. There is a real risk that large portions of the IT industry face a bleak outlook over the next few years, with disruptive consequences for those who fail to deal with the necessary changes that ensue. On the other hand, some pockets will thrive in the midst of the gloom, and companies or individuals who've prepared a strategy for dealing with the worst — if it happens — will have an opportunity to reposition themselves accordingly.

Interestingly, one of Zinsmeister's tips is an area that I know well and which I wrote about earlier this week: "Should demand for web services become mainstream, we think it ultimately should expand the market for IT outsourcing — these web services themselves will be outsourced and managed by third-party providers ... In some respects, web services represent the real market opportunity behind the application service provider (ASP) model."
posted by Phil Wainewright 1:10 AM (GMT) | comments | link

Wither integrators?

It's difficult to see where IT consultants are going to be earning their revenues a few years down the road. Jason Bloomberg recently wrote up some of ZapThink's findings about the impact of SOAs on the consulting market in an article on SearchWebServices. His conclusion: "Over the next five to seven years, the business process design, optimization, and execution consulting market will come to displace the system integration market."

His implication that IT consultants will simply be able to turn to business process consulting is misleading. Business consultants, not IT consultants, will perform business process consulting (as they always have done). If IT consultants want to hold onto their livings when integration work disappears, they'll have to do more than retrain: they'll have to join a completely different profession.
posted by Phil Wainewright 4:18 AM (GMT) | comments | link

The end of software

Consolidation means contraction. When five leading firms propose mergers in the same week, the prognosis for the enterprise software industry looks dire. Especially when in the same week, the most vocal exponent of their nemesis took a bold new step into their universe.

It might have been possible, at a pinch, to explain away two of the mergers. Even though the news was completely unexpected, there's some logic to Peoplesoft's proposed $1.8 billion purchase of JD Edwards, announced last Monday. The two companies have largely complementary products and customer bases, they've both made good progress towards re-architecting their platforms for web services, and they're been working on the deal for many months, as is evident from this excellent account from yesterday's Denver Post. It's a well thought-out, carefully planned merger.

The acquisition of Baan by the owners of SSA, announced the following day, is as logical as it was predictable. It had been public for some time that the fallen former enterprise software giant was up for sale, and the $135 million deal that will see it merged with SSA is a no-brainer.

Each of these mergers, then, can be seen as a calm, measured and logical response to what every observer agrees is some degree of contraction in the enterprise software market. Even though they were announced within a couple dozen hours of each other, the history behind them belies any hint of a panic response to rapidly deteriorating market conditions.

Then Oracle stepped in with a $5.1 billion offer for Peoplesoft, and suddenly two plus two made five. Oracle makes no bones about its motives. It wants to shut down Peoplesoft and grab all its customers. This unresearched, hastily cobbled-together proposal only makes sense as a spoiling tactic, designed to disrupt and derail Peoplesoft's acquisition of JDE. Seen in that light, insiders may still be able to cling to the notion that all these deals are logically consistent with the actions of a robust, thriving industry. But the unmistakeable signal it sends out to anyone looking in from the outside is that the economics of the enterprise software industry just don't add up any more. Every one of these five companies has evidently taken a look at its projected deal flow and revenue expectations, and is preparing itself for a savage drop in demand.

One threat that they don't see coming is personified by a company that launched itself three-and-a-half years ago with a party to celebrate what it billed as "The End of Software." Last Tuesday, salesforce.com announced alliances with leading infrastructure software vendors BEA, Borland, Microsoft and Sun in support of its new Sforce initiative, which allows developers to link salesforce.com's hosted software-based services into their own custom composite applications using standardized web services interfaces. At the same time, the company named 25 third-party software vendors and providers who are ready to plug their own services and tools into the Sforce environment. The proposition is such a sophisticated showcase of rapid-assembly, service-based process automation that few observers had much of a clue what it was all about.

Sforce needs to be surrounded by all the caveats that hedge any early pioneer of a significant new technology trend. Last week, I posted a series of entries to this weblog that amounted to quite a tally of all the holes in the web services architecture as it stands today. There are no mature, agreed specifications for service metrics, billing or reliability, and several of them still seem to be years rather than months away. Until such standards arrive, initiatives like Sforce will remain of interest only to visionaries and early adopters. But that will be enough to build momentum for a new, service-oriented assembly approach to building enterprise software functionality. Today's enterprise software vendors can consolidate all they like, they'll still find themselves unable to contract fast enough in the face of their nemesis.

I would normally write on this theme in a regular weekly column for ASPnews, but it just so happens that my slot came to an end last month — nobody wants to advertise to the ASP market any more, so the publisher has sensibly decided to cut back on its editorial spending. It was ironic, then to see that The Wall Street Journal last week covered the Sforce story under the headline Is renting software online the next big idea? (the link is to an open-access view of the story courtesy of Church Executive Magazine). You can take your pick as to which of these is the contrary indicator. Personally, my take on the ASP phenomenon was that it only ever made sense in the guise of online services like salesforce.com, rather than the more conventional view of ASPs as companies who rented access to other people's software.

The future path of such services is more radical even than Sforce, and to liken it to the notion of renting software online is a bit like imagining at the end of the nineteenth century that the internal combustion engine would be used to power mechanical horses. By the time all of this has played out, software will have changed beyond all recognition, and none of today's enterprise software vendors will survive in their current guise, if any of them survive at all.
posted by Phil Wainewright 11:21 AM (GMT) | comments | link