Can an organization learn?

In response to the last two columns, which talked about the barriers to organizational learning and how to overcome them, a subscriber challenged me on this point. Learning, he said, is something people do, not organizations.

I agree, organizations aren’t just people, only bigger. As has been pointed out in KJR more than once, they’re a different type of creature than human beings, with different motivations and patterns of behavior (see “What corporations and spleens have in common,” Keep the Joint Running, 5/5/2003).

Can organizations learn? As is usually the case, the answer depends on how you define “learn.” Since my scientific training is in sociobiology, I define it operationally: To learn is to change behavior in an adaptive way based on experience.

If that’s the definition, the answer is an unequivocal “yes” — organizations do change their behavior based on what they experience. They learn through changes in the behavior of their executives, managers and staff, just as animals learn through changes in the behavior of individual neurons. The difference: Executives, managers and staff can prevent the organization from learning if it isn’t in their best interests.

To illustrate the point, here is a case study, sent in by a subscriber who really, really needs to remain anonymous:

The company I worked for had several divisions. I was a member of the smallest and we were led by a rogue Product Manager who thought we should write software the clients loved and could understand intuitively.

So we sat with clients, visited their offices to watch them work, etc. Our software won design awards, gained more than 50% market share in its space, and our engineers got standing ovations at user conferences.

At our annual company meeting the CFO would go over “the numbers” broken out by division which was fine when our division was tiny but as we grew people began to realize that we generated several times as much revenue per person as the other division — we had 1/10 the total resources generating 1/3 the quarterly revenue.

Then it got childish. The “numbers” were no longer broken out by division. There was grumbling about special people making more money in “that” division when everyone was working “hard.” The divisional portion of the yearly bonus was discontinued in favor of an overall company performance bonus.

But lo and behold, the success continued. At the peak a division containing 1/10 of all resources serviced 40% of all clients, and generated more than half the new software sales revenue vs. all other divisions combined. That factor of 10 revenue generation ratio was just too much and we soon found our core members redeployed to other projects, new mandates to use code libraries and architecture based on the other divisions’ products, etc.

The moral of the story is that we succeeded for the customer for awhile, as evidenced by our developer-to-customer-support ratio (2 to 1) vs the other divisions (1 to 4) and support-to-customer ratio (1 per 700 vs 1 per 40).

But we failed in the long run because we couldn’t bring the rest of the company along. Our success generated jealousy and animosity in the VP level office suites and since we were so small (1/10 the VP’s) politically we never stood a chance. Once the high and mighty hit on the excuse above we got squashed like a bug.

The company was sold a few years later to a large software accumulator and is now being milked for its support fees. Which as you notice is great when you have a buggy, hard to use product with proprietary data and you charge high support and maintenance fees. But when customers don’t need support because their old version works just fine that model breaks down.

The end result was the mass exodus of the best and brightest and now only a handful of the original group remain. All of us are now seeking to recapture that 5-6 year golden era before we were told to do less well so as not to offend our co-workers.

Truth is definitely weirder than fiction.

If this were fiction — say, an episode of House — we’d put these symptoms on a whiteboard and figure out what disease explains them all. My diagnosis: Weak leadership.

A strong CEO would have created an environment that recognized and emulated success, building it into the corporation’s structure, compensation, and culture. A strong CEO would have recognized that in business, “fair” doesn’t mean equality. It means meritocracy.

This CEO opted for “mediocracy” instead.