Bifurcated Concentration of Knowledge Doesn’t Serve
One assumption behind this model is that the brains are at the top of the organization and the brawn is at the bottom. Knowledge flows down, but mostly on a “need to know” basis. So it filters down as a trickle, not a torrent.
At the same time, the hierarchy acts a filter from the bottom up.
Day-to-day knowledge about how things really work, who to talk to when you want to get something done, how to work the (doesn’t work as defined) process, work arounds, greasing the skids, what the status really is and other useful stuff is concentrated at the bottom of the organization.
The higher in the organization, the more the day-to-day reality of business is abstracted to numbers. And because the people at lower levels in the hierarchy don’t want to displease people at higher levels, they tend to minimize bad news—so the picture becomes rosier with each level it ascends.
One can argue that this model of knowledge concentration has worked. There are many companies that started this way, and they are still around. But I think the days when it served us well are over. It doesn’t meet the needs of the our times or our industry.
For one thing, we aren’t pouring ingots. We do knowledge work. We need to know about the context, customers, and cash flow in order to make the right products.
We live in a time of fast change. The competitive environment and customer preferences can change on a dime.
This bifurcated concentration of knowledge does not serve us. It makes companies slow as requests for approval work their way up the chain. Worse, it makes companies stupid. Think about it. How many times has a decision from top management looked silly, misguided, or malicious given the realities of implementation? How many times has management countermanded a decision made close to the work because the people who made the decision were missing some piece of information about market context, customer needs, or how cash comes in the company door?
If we want our companies to succeed, we need to make the over-lap between contextual knowledge and implementation knowledge bigger.