11 August 2004 When it comes to implementing new technologies, businesses don’t like to act in isolation; they look for validation from peers and analysts.
That is what makes the latest batch of research reports concerning web services so confusing. According to one set of research, the technology is overwhelmingly popular with enterprises. But according to another, 40% of enterprises have no plans to use web services at all.
Only when the customer looks behind the headlines and more closely at the latest figures and statistics does it becomes clear that, in spite of some wariness, web services is now genuinely becoming a mainstream technology.
The chief web services cheerleaders are the young US technology advisory firm Nemertes Research and the long established research group, the Yankee Group.
Nemertes says that 75% of the IT using organisations it surveyed are using web services to communicate with trading partners, customers and suppliers. And the Yankee Group, known for its bullish forecasts, says that 70% of new IT spending is on deploying standards-based integration technologies to connect the enterprise with those groups, too.
The main sceptic is AMR Research, a company with more extensive and deeper ties to its clients in the user community: its research suggests 40% of organisations have not deployed web services in their organisation and have no plans to do so. And of those that have, 73% have fewer than five web services projects in production.
Caught in the middle is the giant thought-leader, the Gartner Group, which says that sales of web services-enabled software products and related services will grow from $61 billion to $316 billion between 2003 and 2007.
Clearly, they cannot all be right. Or can they?
A look at the small print shows that there is less difference between them than at first sight. Nemertes’ research is based on a survey of 35 medium and large enterprises with annual revenues of over $500 million – too small and skewed a sample to give a true idea of web services’ popularity.
The Yankee and Gartner research focuses on web services spending. And this in itself can be misleading. For example, Gartner doesn’t specify whether the strong upsurge in web services software expenditure is the result of interest in web services per se, or a by-product of the increasing web services enablement of enterprise software that these organisations might be buying anyway. Today, all middleware platforms increasingly use web services.
Most intriguingly, Gartner’s research suggests that the majority of the spending ($271 billion) will be on web services consultants. In conjunction with Yankee’s research (which only specifies that a high proportion of the total IT spend by all organisations will be on integration technologies), this suggests that web services technology is not the cheap integration tool initially hoped for. As with previous innovations, users are going to be offered, and will presumably buy, a lot of handholding and help.
Unfortunately for those more zealous web services advocates, and importantly for those treading warily and watching watch everyone else is doing, AMR’s sceptical approach looks justified. Its research, moreover, is based on a large and broad sample size for it to be valid.
AMR’s position is not that web services is not going to be important, but that the costs and complexities will be greater than many people have realised and that adoption will be slower than has been widely forecast.
Taken together, these analysts’ research paints a clear picture: while web services is becoming more popular, it is also proving to be more expensive than initially marketed; take up is still in its earlier stages.
A secondary finding is that web services will initially be of much greater use to larger organisations than to their smaller counterparts, not least because of the complexities.
Web services is not matching up to its initial hype – how could it? But with even AMR’s sceptical position suggesting that 60% of organisations are already implementing the technology, it is also clear that web services is not already mainstream, then it very soon will be.