.comment-link {margin-left:.6em;}

Thursday, December 15, 2005

Improving the function of lessons databases

Organizations need to gather and utilize past "lessons learned" but too often, critical information is lost or simply not used. Here, Alan Thompson of KBR Production Services gives five top tips for improving lessons databases.

1. Ensure your lessons database has a balance of successful and unsuccessful lessons. A narrow "blunder avoidance" database full of, "don't go there - we did and it failed" examples, affects the way personnel perceive lessons and also affects the organization's clients' perception of processes.

2. Make the context, categorization and specific details of lessons clear if you want them to be used appropriately. Improve the categorization of a lesson by asking a lesson user to check if the lesson is valid, current, in the correct context and available and for use.

3. Develop an expert or skill finder system for your database to give users a choice of knowledgeable contacts. Include relevant information on experts so the user can make a specific choice and quickly get the information they need. Make experts "duty bound" to answer callers to maintain the function of the system.

4. For organizations with knowledge repositories spread across several countries consider the use of a "loose federation" database model, where each business unit "owns" their own data and the processes from which it was derived. In this model, users primarily look for lessons from their own business unit, then spread the search out across the company to other business units who in turn, decide if they can help with their own knowledge and "broker" the information they have.

5. Ensure your database remains fresh and useable by using it yourself with your own projects.

Adapted from "Getting real value from lessons databases" by Alan Thompson in the current issue of Knowledge Management review.

~~~~~ .. ~~~~~

Closing the loop on knowledge re-use

Putting a knowledge sharing process in place is only half the challenge to improving knowledge sharing across an organization. You must also ensure that published knowledge is used and replication processes are acknowledged and implemented by business units (BUs) that need them most.

In order for replication and knowledge to be implemented on poor performing BU's it might be necessary to make mandatory requirements of implementation.

For example, if a top performing BU submits a knowledge document, a best practice process for example, it can be mandatory, not optional, for BUs that were not in the top performing band for the relevant business measure to complete replication and report the results.

In a small minority of cases, BUs may report after studying published knowledge submissions that they either have similar practices or cannot replicate due to specific local reasons. Such genuine exceptions should be accepted, but it's should be mandatory for the BU to close-loop (respond to) every single published knowledge submission.

The status of replications and results can be made part of the agenda for the president's review with CEOs and each CEO's review with their functional heads. Particular attention should be paid to BUs in the bottom performing bands for each business measure (over 25 percent variation from the best performing BU) to ensure that they complete replication of best practices in those areas where their current performance is weak.

The knowledge replication process should not end with a BU completing replication of a knowledge submission relevant to them. Make it mandatory to measure the quantified impact of the replication on the relevant business measure.

Adapted from "Establishing KM Processes Part Two" by Arun Hariharan in the current issue of KM Review.

~~~~~ .. ~~~~~