A KM star is born

This narrative was told to Jerry Ash during interviews for the case report on Guerilla KM. It is presented here in Gent’s own words.

By Andrew Gent and Jerry Ash

“FOR AS long as I’ve known him, Stan Garfield has advocated a points system for encouraging KM activities. But we have never had the time or resources to implement such a system. We saw some third-party products along this line, but they involved a closed system of KM tools, and in our environment we knew we had to address multiple heterogeneous systems. So the challenge was, could we build one ourselves? After some discussions, some detailed design work, and back and forth, Stan finally asked what could be built in two weeks.

So we stripped the design down to its basics: a database, an entry form, a frequently asked questions (FAQ) page, and reports (perhaps the most important part, as the whole goal is to encourage participation through recognition). We took out all the automation of points so users must report on their own activities. And ‘KM Stars’ was born.

We also made two other important design decisions. The first was to make the system opt-in; users must register before their scores are included in the reports. This avoids any privacy concerns (important in Europe) and reinforces the ‘only if you want to’ aspect – there is no compulsion.

Of course, we give users an extra five points just for registering. The second important design decision was to limit the system to recognition; we decided not to associate any rewards with the points. This, again, simplified the design. Having been involved in other reward programmes, I knew it would take months to get approval and to work through the details of anything involving payment of cash or gifts to employees. So, again, we offered KM Stars as a method of measuring individuals’ KM activity, but left it to the other organisations (which have the money) to grant awards based on KM Stars as they saw fit.

The response was as might be expected. There was interest, but many more concerns that users wouldn’t use the system if they had to report their own activities. It needed to be automated, which was true enough. Initial adoption was slow (around 800 individuals). But three things changed that situation:

First, the reports. Remember I told you I am big on RSS [really simple syndication]? Well, the initial reports were written as RSS feeds. As a consequence, they are very easy to include in web pages. So we included the weekly top ten (reports are available as weekly, monthly, quarterly or lifetime views) on the KM and other web sites;
After a few months, I finally found some more spare time to write a scheduled script to automate points for postings into the forums. Again, this is done by reading and interpreting an RSS feed from the forums;
Finally, one of our advocates in Europe convinced her organisation to grant a small award to the top point-scorers in their organisation.

This last item was a major win: as an incentive for people to participate, as an advertisement of the system and as a source of the next major upgrade. To issue the awards, they had to read through the reports looking up each name to find their group members.

So the second major update to KM Stars was to build a monthly report that lets you sort/filter the reports by country, region, and reporting organisation.

Again, this was written in about three days. It is monthly because that was the quickest way to do it – it actually extracts the points and builds a separate database. This avoided serious testing and debugging if I tried to modify the existing database.

Now, several organisations are using KM Stars to grant awards. Registration is still low-ish (2,000+), but growing. And all with little or no financial investment.

KM Stars also demonstrates a few of the other characteristics of guerrilla KM projects, both good and bad. We receive many suggestions for new features. Perhaps the most frequent from management is a ‘detailed’ report listing what an individual got points for. One organisation has stated it won’t use KM Stars as a basis for rewards because the organisation cannot ‘trust’ it and only wants to grant awards for some of the activities for which points are granted.

Normally, all of these features would have to be considered in the original design. However, one of our basic design principles for the system is to recognise the cumulative effect from small day-to-day activities, not detailed reporting. So we have rejected this request. This causes some grumbling from management, but having a clear design goal is essential to keeping the system clean and simple. Besides, if they want to grant awards for specific activities, there are plenty of ways for them to count them themselves, if they want to spend the time.

Also, because of the preference for fast over perfect, sometimes seemingly simple changes can be almost impossible because of design decisions made to save time. To get this done in two weeks, no effort was spent on administrative interfaces, for example. So any maintenance had to be done through database hacking. Not good.

We have limped along for a year-and-a-half this way. But I finally broke down and spent the last three days building an administrative interface to make management easier.”

Andrew Gent is lead knowledge architect for HP Services Consulting & Integration. He lives and works in Nashua, New Hampshire and can be found at http://incrediblydull.blogspot.com/. Andrew will be guest moderator for AOK’s STAR Series Dialogues, 17-28 September. AOK membership is free and you can join online at: www.kwork.org/explain_join.html. You can contact Gent directly by e-mailing andrew.gent@hp.com.