Don’t fall for that false dichotomy! Codification vs. personalization

One of the points that has long been made about knowledge management is that most KM initiatives can usefully be categorized as either focusing on the codification of information and knowledge or on the personalization of information and knowledge. IBM (ibm.com), for example, has for several years--that is from very early on in the growth of KM--employed the chart on this page to help explain the domain of KM:

80-20

The horizontal dimension is precisely codification: collecting (stuff), and personalization, i.e. connecting (people).

However, in a recent article entitled “What’s Your Strategy for Managing Knowledge?” in Harvard Business Review, Hansen, Noheria and Tierney have seized on the codification/personalization distinction to argue that not only are most specific KM undertakings oriented to either codification or personalization , but that overall organizational strategies for KM are similarly either heavily codification or heavily personalization. They argue that knowledge management systems are almost inevitably an 80-20 balance, with some firms emphasizing the codification--for example, Andersen Consulting (now Accenture and Dell --while others--such as McKinsey and Hewlett-Packard --emphasize the personalization. They point out that a firm must think carefully about its knowledge management strategy and align its strategy with its business operations and goals: codification if an organization’s products are standardized and mature and if people rely primarily on explicit knowledge, and personalization if an organization’s products are customized and innovative, and if people rely primarily on tacit knowledge.

Their points are pretty straightforward and relatively unexceptionable to this point. But then, however, they go on to argue an additional point that not only are knowledge management strategies almost always fish or fowl, either 80% codification and 20% personalization or the reverse, but that straddles, something like a 50/50 mix, signal a lack of focus strategy and are to be avoided. Their advice is very specific; “DO NOT STRADDLE (their capitalization).

50-50?While the advice to match one’s strategy with the context is admirable, the “Do not straddle” advice is overly simplistic; indeed, it is dangerously misleading. An analysis of the pharmaceutical industry, for example, reveals a situation where a 50/50 straddle is precisely what is called for. In my studies of the relationship between the research productivity of pharmaceutical companies and the information/knowledge characteristics of their research environments, I found a fascinating phenomenon. In the pharmaceutical industry, a heavy emphasis on codification is, of course, required. Codified access to research notebooks, compounds synthesized, screening results, clinical trials, etc. is de rigueur. The research revealed, however, that the most salient difference between the less successful vs. the more successful pharmaceutical companies is that the less successful firms have a knowledge environment with an 80-20 emphasis (on codification), while the more successful companies , have an equal emphasis on codification and on personalization. They have deliberately adopted the 50/50 straddle that Hansen, Nohria and Tierney advise us to avoid. And research on the pharmaceutical companies is not anecdotal; it is quantifiable and rigorous.

In conclusion, companies should think carefully about knowledge management strategy and align that strategy with their business operations and goals. However, the 80-20 and 20-80 emphases are not either/or choices; they are the practical limits of the range in which the appropriate knowledge management strategy mix lies. The correct balance for overall KM implementation may lie anywhere within that 80-20 to 20-80 range, and further, it is likely to differ within different functions or units of the organization. Don’t fall for the “Do not straddle” advice.

Dr. Michael Koenig is dean of Long Island University’s Palmer School of Information Science, e-mail mike.koenig.liu.edu