[KM-Forum] CoP vs KM Portal = ROI?

Patrick Lambe Straits Knowledge SG plambe at straitsknowledge.com
Wed Aug 30 11:28:02 IST 2006

The problem with ROI in KM is that strict ROI requires the ability to
demonstrate direct causal linkages between input (investment) and output
(outcomes). That's very hard to demonstrate in many cases, however hard you
try, because

(a) KM is usually infrastructural, and infrastructure is a sunk cost that
enables or support more directly attributable operations (such as sales
activity or specific functional IT applications) that tend to take the ROI
credit - as a parallel, economists still have no uncontroversial models for
calculating the ROI on national infrastructure developments, they believe
strongly it has a positive impact on national outputs, but they don't know
how to calculate how much. John Chambers of Cisco once famously said, when
asked if he had quantified the ROI for the use of elearning in his company,
"What's the point? How would you calculate the ROI on your telephone system?
We just know we can't do business without it."
(b) KM is usually multi-strand, diffuse, and complex, and so even the
investments, let alone the outcomes, are difficult to measure and quantify.

I agree that where resources are allocated there is a fiduciary
responsibility to try to achieve maximum value for the investment, and
therefore the evaluation question cannot be avoided. The trend in this area
(and in related areas such as investments for community development
projects, national knowledge infrastructure, ecological interventions,
national reconstruction etc etc) seems to be to move towards a concept of
"value" creation rather than a strictly monetary input-output model. So
measurement models are not strictly quantitative any more, in fact KM
standards bodies seems to be suggesting a "portfolio" approach to
measurement where qualitative assessments of impact are gathered and linked
wherever possible to quantitative measures.

Good examples of this can be found in the British Library's adoption of the
contingent valuation methodology www.bl.uk/about/valueconf/pdf/value.pdf  or
the development of the Most Significant Change methodology by development
agencies http://www.mande.co.uk/docs/MSCGuide.htm

Everyone who's worked in this field seems to agree that strict financial ROI
is too simplistic to be able to model the complex ways in which
knowledge-oriented initiatives work their way through our infrastructure.
The Australian KM Standards document AS5037:2005 as well as the 2003 report
by APQC on "Measuring the Impact of Knowledge Management" both recommend
hybrid, portfolio approaches.

Of course, there will be exceptions, where a KM initiative is very tightly
contained, the investment is easily quantified, and the goals are very well
defined, measurable and precise (eg putting in a collaboration platform for
a sales force in the absence of any other performance or productivity
improvement interventions). But such controlled conditions are very rare.

Best wishes


Patrick Lambe

More information about the KM-Forum mailing list