Knowledge Generation

Abstract

In the previous chapter, we presented a systematic way of using the firm’s knowledge to improve the service rendered to the customer. In this particular case, it was a procedure in the form of an improvement project - that is, a one-off event within the firm’s life. But how can this approach be transformed into a permanent improvement system? At first sight, the answer might be: ‘Well, let’s carry out the method once a day.’ However, when put into practice in this manner, the process suffers from two basic drawbacks:

• 1. The preparation is costly. Any knowledge identification process, discovery of moments of truth, and the like, that is needed to arrive at a list of actions, is time-consuming. To perform it every day is costly. But, on top of that, it is redundant. Because, unless drastic changes are made within the firm, we will discover the same knowledge that we had the previous day. And in the process, we will perform that wondrous act of carrying out a costly process that contributes nothing new. This is logical, because knowledge is only changed by our actions and subsequent learning. However, in one day, not enough time will have passed to perform actions that change the knowledge base. One possible modification is to lengthen the time between analyses. However, now the opposite effect happens. If we lengthen it too much, we will miss opportunities because the knowledge is not detected when it is generated.

• 2. It is a system of plans. The outcome of the process is an action plan consisting of a sequence of actions that encompass a not insignificant time period. The plan represents a commitment to future action by the individuals involved, sanctioned by the firm’s management. This has three effects. First, it gives the plan an internal application logic, with precedents and synergies between activities. Generally speaking, it is difficult to separate the plan into individual actions that can be performed concurrently by the agents and which can be generated by each agent in the course of interacting with his daily work environment. Second, the plan must be reviewed by consensus and iteration. With this method, it is unlikely that each agent will have the initiative to change actions as problems are detected. The result of the analysis is both too detailed - with too many coordinated actions - and too broad to enable rapid review of the plans. And lastly, in order to assess actions, they must all be available and have been described in sufficient detail.