KM Archive Home   Site Guide   I3Update   Insights   Resources   K-Shop   About Us   Contact Us     ::Our new (2011) KM website is kmknowledge.com ::
 

 

January 2003    Feature
a free monthly briefing on the knowledge agenda
 
 
No. 69
Cover

Main Feature

Other Features





Latest Edition

Archive

Knowledge
Digest


About I3

Feedback



Managing editor:
David J. Skyrme

Publishers:


Models for Metrics:
Are Yours Mechanical or Mental?

David J. Skyrme

The last few years have seen a growing interest in measurement methods of intellectual capital. After all, if a company's balance sheet only tells a partial story - and in some cases a misleading one at that - then we need to focus on measures that are related to a firm's true value and factors that drive its long-term sustainability. These factors include innovation, knowledge, governance, quality of management and leadership, specialist know-how and so on. The problem is that most of these factors are intangible, whose measurement lack precision and is therefore abhorred by accountants. However, ignoring what is required for legal accounting purposes, models that focus on these drivers are receiving considerable attention.

Why measure intangibles such as intellectual capital?

There is growing evidence that indicators of intellectual capital are leading indicators of future financial wealth. Early evidence came from a study of advertising agency Saatchi and Saatchi by Karl-Erik Sveiby and Tom Lloyd in 1987. Even though its stock value was rising, many indicators of IC were falling, such as number of creative staff, number of key accounts, quality of customer relationships. Sure enough, a few years later, Saatchi and Saatchi's finances deteriorated significantly. More recently, studies of innovation, human resources management performance and quality of corporate governance (to mention but a few) have shown significant correlation between the effectiveness of these and medium-term financial performance. A pilot project initiated by Denmark's Ministry of Trade and Industry in 1997 encouraged 19 companies to generate IC reports (today around 100 Danish organizations are involved in such an exercise). Typical of the benefits cited (see Danish Guidelines( (PDF file)are:

  • Better communication with stakeholders, not just shareholders but customers, suppliers, business partners and employees.
  • Better image of the company, showing that it understands its drivers of wealth.
  • Insights into the drivers of wealth and the interactions between various components of intellectual capital.
  • More attractive company to prospective employees (since human resource indicators show commitment to training etc.)
  • Better resourcing and investment decisions, concentrating resources on value drivers.
  • Improved integration of performance measurement into daily activities.
  • Higher automation of indicator data collection.

Top-Down, Bottom-Up and Knowledge In-Out

There are several different approaches to measuring intellectual capital. The first category includes those that take an organization's reported figures and adjust them to remove 'anomalies' of traditional accounting. One of the most familiar is EVATM (Economic Value-Added). Those taking a knowledge lens include human resource accounting and several variants of Knowledge Accounting (Benjamin) and Knowledge Capital (two different methods with the same name by Lev and Strassmann). Rather than regarding people as costs, a significant proportion of employee costs are viewed as assets. Similar considerations apply to R&D expenditure. These methods are top down, and though they can quickly give you an estimate of intangible value (and value-added) in &# terms, give relatively little insights into how it is created and developed.

The second category of method is more bottom-up. These methods - the Intangible Assets Monitor (Sveiby) is an example - consider the various categories of intellectual capital, from which are developed various measures. A typical categorization of intellectual capital is human capital (this is universal), structural capital (whose scope and definition varies, but as a minimum includes an organization's processes and systems) and relationship capital (sometimes the third category is customer capital, but the term relationship capital is wider in scope and recognizes the value generated by other stakeholders). The most tangible form of intellectual capital - intellectual property - is the one that no two models categorize in the same way! Early IC models simply attempted to develop indicators for each category of IC.

The third category of method, is some form of scorecard. These don't explicitly identify intellectual capital, though in most scorecards there are categories that obviously include it, such as the innovation and learning dimension of the balanced business scorecard. Most of these methods start from the corporate strategy, then evolve a balanced set of indicators - financial and non-financial - that can be tailored to various levels within the organization e.g. business unit, team, even individual.

The final category of IC method (in this article we're not covering various KM assessment and ROI methods), comprises "third generation" IC methods. These take account of the knowledge flows between different forms of IC (as in IC IndexTM). Thus, as well as indicators of stocks (how much expertise do your people have), they consider activities that help develop it (e.g. number of training days) and conversion activities (e.g., capturing best practices from people's tacit knowledge). Additional refinements are visualization of results (e.g. QPR scorecard) and addition of risk ratings (IC Rating) and computer software to perform "what-if" analyses (IVM).

The bottom line of such methods are indicators that show how well you are addressing the different aspects of IC value creation, what knowledge recipes you apply to the knowledge you have, and what the outcomes are (that's why we've labelled it knowledge in-out).

There are variations on these themes and the methods are getting more sophisticated all the time (some too sophisticated for practical day-to-day usage). In fact, our forthcoming update to the report Measuring the Value of Knowledge we compare and contrast over 30 different methods.

Numbers and Narrative

One of the difficulties of measuring intangibles is that any numbers associated with them do not follow traditional mathematical rules. Thus 2 + 2 can equal 6 or even 60, if the combination of two disparate types of knowledge leads to breakthrough innovations. On the other hand, as we know from many merger situations, 1 + 1 can lead to a haemorrhaging of talent and a combined value less than the sum of the parts. Similarly, many individual components of intellectual capital are difficult to measure precisely. How do you measure the value of hands-on (or rather heads-on) experience? Much of IC measurement is subjective.

There's an analogy here to the codification of knowledge. What starts off as ideas in people's heads is not easy to grasp. Its only as it starts to become articulated in some more concrete form - an engineering drawing, a description of a method, a piece of software, and so on - that it there is some commonality of understanding and precision. Likewise with IC. With many aspects difficult to conceptualize, a raw number that is a measure of an indicator conveys little to many people. However, couple it to some narrative about the concept and how such and such knowledge helps organizational performance and people's understanding increases. As it does so, it is then possible to look for surrogate indicators. For example, a measure of experience might be how quickly somebody solves a particular type of problem. But its only after measuring this over a period of time, and identifying other factors that might improve problem solving that will tell how good such an indicator is.

In other words, the narrative behind the numbers is what increases our understanding about measures.

Correlations, Cause and Effect

Behind all IC measurement methods is some kind of model. At one level it may simply be a hierarchy of types of IC. However, such a basic model does not help explain how to improve the output measures. By applying statistical methods we might start to discern correlations. We might, for example find that there is no correlation between the amount of a person's training and their experience (using the indicator suggested above are not correlated). However, it might be that the time in a given job has a degree of correlation (don't we all like dealing with account representatives who have after our assiduous tutoring have just got to know our requirements but then get promoted to management!). A not untypical correlation is that between employee motivation and customer satisfaction.

But correlations don't immediately tell you why something happens. You may have a hypothesis. So you could build a systems model from your hypothesis and test it out over a series of data over time. That's what economists and weather forecaster do, but our experience of their results shows that methods take a long time to perfect - if they ever do! Contrast explicit modelling with data mining techniques, that might indicate that sales of beer on Fridays increase when placed next to baby's diapers. You don't need to figure out the reason why (like many of these associations, they may be obvious in hindsight), you just put one and the other together and watch your sales climb. On the other hand, it is satisfying to know why, since that may spark off ideas for other innovative knowledge recipes.

Are Your Models Mental or Mechanical?

The bottom line of measurement is that you are measuring for a purpose. If you treat IC measurement as simply a numbers game, you lose a lot of richness. Virtually every user of IC measures - or other performance measures come to that - say that talking about the results and discussing differences with their peers can give powerful insights into how their organization, and their parts of it, work. Thus, benchmarking, for example, is rapidly becoming overtaken by the term benchlearning. You measure performance in different organizations and dialogue the differences. That way you learn. That's much more beneficial to long-term value creation than simply collating a set of numbers for accounting purposes.

So, instead of models that you populate mechanically, think of your IC measurement activities as a way to inform your mental models and truly create value for your organization and yourself.


I3 Update / Entovation International News:
Current Issue - Archive - Knowledge Digest - Events - Subscribe


Top of Page
Home | Search | I3 UPDATE | Insights | Resources | About Us | E-mail Us

INTELLIGENCE INSIGHT INNOVATION