Experimental feature

Listen to this article

00:00
00:00
Experimental feature
or

Fans of knowledge management see it as a potent tool for transforming a company’s most important asset. But critics argue knowledge management is an oxymoron.

Information, by contrast, seems more manageable and tangible. So vendors of information management tools such as business intelligence (BI) are given a warmer reception than those peddling knowledge management.

The IT industry has traditionally presented BI and KM as complementary solutions to a common problem: how to make sense of the mass of information held by any modern business.

“BI is about analysing data while KM is about sharing content but both approaches are complementary and they are not dependent on one another,” says Shari Rogalski, global executive director for BI at Accenture.

BI software extracts data from databases and turns them into human-readable reports. KM, by contrast, has its roots in linguistics and aims to help organisations share and apply the “hidden” knowledge of employees.

However, the boundaries between BI and KM have become blurred over the past decade and neither has lived up to early expectations. This is certainly the case for KM, which has suffered the hype typical of management fads.

In spite of the emergence of “social” technologies such as wikis and weblogs, most organisations lack the culture and processes needed to promote the organising and sharing of knowledge. Experts say KM initiatives that do not address these are doomed to fail – as many pioneers have discovered to their cost.

Another problem is coming up with a workable definition of KM – and knowledge itself. Most people would accept that knowledge is different from information, but in what way exactly?

Software vendors skirt round this philosophical challenge and typically present knowledge as a more valuable type of information. That has led them to attach the KM label to a bewildering range of software, including portals, document management systems, query and reporting tools and online analytical processing (Olap) systems. To add confusion, many of these technologies also feature in BI offerings.

SAP, the leading enterprise software company, prefers not to take sides and claims its NetWeaver technology supports both camps. Richard Brown, head of BI for SAP, says some customers buy NetWeaver to use as a standalone BI engine, using features such as data mining, alerts and pattern matching.

Others prefer to put NetWeaver at the centre of a more ambitious knowledge-sharing initiative using the enterprise portal component of NetWeaver and its enterprise search capabilities. “Some customers want to go beyond BI and get the bigger picture using contextual information,” says Mr Brown.

According to Lee Phillips, director of knowledge solutions at FAST, the Norwegian search engine specialist, the key to understanding the difference between BI and KM lies not so much in the underlying technologies but in the problem they are trying to solve.

BI tools concentrate on analysing structured information held in databases – personnel records, regional sales figures and so on. KM, by contrast, tries to capture and make sense of unstructured information that lies outside databases, such as emails, slides, web pages and so on.

“Organisations want to put all their unstructured information on the same footing as the data they analyse using BI,” says Mr Phillips.

One of the fastest growing types of unstructured information is emails. As financial regulators discovered after the dotcom crash, emails can provide a far more accurate picture of what is really going on inside a business.

Wall Street regulators seized private emails that revealed investment banks knew their analysts were writing bullish reports on dotcom stocks they privately felt to be unsound. The scandal cost the companies concerned $1.4bn.

“Unstructured information is every bit as important as structured information but the difference is that humans are not good at analysing it because of the natural ambiguity of language,” says Mr Phillips.

Business language is usually less forthright and meanings can be highly ambiguous. So the hard work in KM lies in definitions.

“There is a lot of work in getting everyone on board because the whole concept is kind of fuzzy,” says Ms Rogalski.

Experts say KM can work well if everyone agrees on the semantics and the task is not too complex. But it is clear KM has not lived up to its promise and the technology has not developed as rapidly as BI. “BI has a lot more traction than KM because it addresses a very concrete business problem,” says Mr Brown.

This lead looks set to grow as BI vendors such as Business Objects and SAS Institute add capabilities that work with unstructured data.

In the longer term, Ms Rogaski believes the disciplines of BI and KM will converge and disappear as mainstream vendors such as Oracle, Microsoft and SAP snap up BI, KM and enterprise search specialists to incorporate their technologies into their own software suites.

Copyright The Financial Times Limited 2017. All rights reserved.
myFT

Follow the topics mentioned in this article

Comments have not been enabled for this article.