Knowledge management (KM) is the market that dare not speak its name. During the mid-1990s, when organisations realised that they had far more unstructured data in their systems

than anyone could understand or locate – and even more held inside the heads of their employees – KM became a buzzword that just about every self-respecting, content-centric software vendor had to mention. But as IT budgets came under pressure, many organisations began to question if the often-intangible benefits offered by KM were enough to justify the cost and administration overheads. Many stopped their implementations and set their ‘knowledge officers’ to work elsewhere.

As such, the KM term has begun to disappear. Yet the challenge it sought to address – the capturing and cataloguing of the core information generated by the organisation’s ‘knowledge workers’ and the task of making that available to others – did not go away. It just became aligned with a wider set of requirements.

“Very few vendors today position themselves as knowledge management providers,” says Brenda Morris, UK managing director at content management and KM software vendor Open Text. “Instead, you increasingly hear people talking about things like collaboration, content management and business process management.”

Some kind of ‘knowledge management system’ may be wrapped in with those, but fewer organisations are demanding a standalone KM product, with the possible exception of so-called knowledge-rich companies, such as law firms and consultancies.

For organisations where the main business is not necessarily knowledge, the pay-offs from a KM system are less clear, say critics. Being able to mine for information within these organisations is still an appealing notion, but the difficulties and cost associated with capturing employees’ knowledge has made the technology less than compelling.

‘Goofballs’

The main problem is that KM cannot achieve all its aims without changes to an organisation’s culture and business processes. Only if a company is ready to adopt an approach that encourages brain-dumping and knowledge-sharing between employees will a KM system be able to flourish, says Plumtree’s VP of product management, Glenn Kelman. Otherwise, only information gathered directly by computer systems can be shared – although that content in itself can constitute a knowledge database. “You get goofballs talking about how the tacit knowledge in people’s brains forms part of a knowledge management strategy,” he says. “But obviously you can’t write software to capture that stuff – short of drilling a hole in their heads and hooking it up to electrodes.”

The upshot is that vendors are focusing more on an amalgam of content and knowledge products when selling to enterprise customers. While many existing users that have seen benefit from KM are moving to bigger and wider implementations, the primary focus of new customers is on related technologies – such as portals, search engines and categorisation systems.

Demand in some of those segments is significant, particularly at the departmental level. Mark Kenyon, European regional director of IBM’s Lotus division, says that while sales of KM systems are growing at a steady single-digit rate, sales of portal software shot up almost three-fold in the first half of 2003.

“Customers are much more savvy these days,” says Ian Black, a director at information management software specialist Autonomy. “They’re more likely to say, ‘Here’s the [knowledge] problem I’ve got, if you can help me, prove that it delivers benefit and I’ll buy’.”

The requirement for a broader software base has resulted in a flurry of acquisitions by market leading suppliers.

Over 2003, content management software vendor Documentum bought web collaboration specialist eRooms (and then more recently was itself acquired by storage systems giant EMC); Vignette, also from content management, acquired collaborative content software vendor Intraspect; and Open Text took over web content management and portal company Gauss Interprise, portal company Corechange, and document management and sharing software company Ixos.

Many companies are also trying to refine their product’s search and categorisation technologies to match the kind of capabilities displayed by Internet search engine such as Google. But they have some barriers to scale. “Google has the advantage of being able to pick from three billion documents,” says Alkis Papadolpoullos, director of linguistics at Convera, a search and KM company. “It’s likely to produce something relevant for you, provided you haven’t searched for something too generic. But you still can’t tell if what Google has produced is the best result for you. On a corporate network, you have far fewer documents to pick from and it’s more obvious to you if the result isn’t what you want. You need to be able to give the system a sense of what documents are about through taxonomies to produce good results on an internal network.”

Taxonomies, both enterprise-specific and personalised, will be one of the key ideas for KM systems in the next few years, says Edward Truch, director of The Knowledge Management Institute. And depending on its application, enterprises will either want vendors to develop specific categories or they will want to do it themselves.

While the KM label may have lost its lustre, subsumed into what Gartner analysts call ‘smart enterprise suites’ or broken down into components applicable to organisations’ different needs, the underlying requirements for the technology are still growing. KM isn’t dead; it has just been subsumed into something bigger.