Michael: I think you may be worrying about the theory, and ignoring
the practical results. As the economist says, "Well, that's all
well and good in practice...but will it work in *theory*?!?"
"Knowledge databases" are cropping up all over inside corpora-
tions we work with, although they are hardly compliant with the
theoretical models you postulate. I fear the problem lies in the
definition...it certainly doesn't lie in the results.
There are three common classes of "knowledge databases" that have
experienced immense success, in my opinion:
1. Contact and account management, where individuals post their
individual experiences with contacts in customer (or vendor)
organizations. The collective view one can get of what's
going on in an account is rich and varied.
Lewis Burns, of Intel, talks publicly about a situation where
a field sales user of such a database posted an interesting
item about what a competitor had told a client. A few days
later, another sales rep reported something similar from an-
other continent. Finally, when the third example arrived,
sales management recognized a pattern ("One's an event, two's
a trend, three's a pattern!". As Lewis asks, "What's the
economic advantage in knowing your competitor's change in
strategy in a few days, compared with waiting until the
business is lost and you have to try to recover it?"
2. Shared best practices, where individuals post their various
trials and successes (although we try to encourage them to
post their failures, too...but with little success). This
is particularly useful in multi-national manufacturing con-
cerns, although as some have mentioned in this thread,
the "Big Six" accounting and consulting firms tend to use
it, too.
The problem here is linking the system with the compensation
plan. In "Big Six" firms, the route to financial success is
through promotion to Partner. You get to be Partner based on
your annual billings. So, if I'm a CPA in Chicago, and I've
found a particularly clever solution to a tax problem, it is
NOT in my self-interest to post that solution to a knowledge
database. What Wanda Orlikowski reports in her paper on
the pseudonomynous "Alpha Corp." is that our Chicago CPA will
post *that* he's found a solution. Then, the CPA in Los
Angeles will have to bring in the Chicago expert for delivery,
thus putting the billing bucks on the Chicago ledger.
3. Organizational learning models, where collaborative exper-
ience is shared to yield progressively better results.
In a major experience we've helped foster here, one client
invested about $400K in a pilot program (and will invest an-
other $500K this coming year) to yield a $10 Million benefit
*per annum*. The company builds its own factories. Big,
expensive factories, that cost upwards of $2 Billion, and
take 18 months to complete. When they're done, factories
produce about $1 Million/day in gross margin. By gathering
information from all the people who are involved in design-
ing the processes, and the new materials, and the buildings
themselves, and adding that to data gathered from the people
who actually pour the concrete and erect the walls, the firm
is getting measurably better at producing higher-quality
buildings, cheaper. And, they're doing it ten days sooner.
Chopping ten days of 18 months isn't hard...but at $1 Million/
day in gross margin, it has enormous dividends.
We operate from a less theoretical model that you, Michael. We
use a four-level scale:
Data "7"
Information "Number of days in a week = 7"
Knowledge "We have 7 days to complete this task"
Wisdom Knowing whether you CAN or not
So, your concerns about the *potential* for companies wasting a
lot of money is there. But the results suggest that the benefits
are often achieved. In a recent study we did for a client (which
we'll be summarizing at a tutorial at Groupware '95 in San Jose)
says that these projects have about a 50-50 likelihood of success.
The critical success factor is whether you attend to the culture
of the using organization or not (In #3, above, we didn't at-
tend *enough* to the general contractor's culture...which we're
about to remedy with the new-and-improved system we're rolling
out next year.)
Incidentally, most of these projects that we know of that have
been successful have been implemented in Lotus Notes (which I
also happen to be using to write this memo).
-----------------------------------------------------------------
Carol Anne Ogdin "Great minds discuss ideas,
Deep Woods Technology average minds discuss events,
CAOgdin@deepwoods.com small minds discuss people."
--Adm. Hyman G. Rickover