Monday, May 24, 2010

Organizational Thinking

I read somewhere that Google is apparently looking for a head of social. I don't know how much of the press around this is real, and whether the reports on the experience were from highly qualified candidates who were treated badly. My book, An Engineer's Guide to Silicon Valley Startups described the situation when I joined Google, which was that the engineering organization not only couldn't figure out how to hire release engineers, but didn't know what it didn't know about how to hire appropriate release engineers, leading to a logjam and deadlock (not to mention some unhappy auditors).

When I gave my talk at LOLapps.com and mentioned this, someone from the audience asked me this insightful question: "How do you learn to recognize these deadlock situations? How do you hire or train someone to recognize these types of situation?" My initial response was that if the founder/executive team doesn't recognize these, you have a problem. My next response was to point him at one of my favorite organizational analysis textbooks: The Fifth Discipline. One of my favorite chapters of that book is the one on The Beer Game. It shows how given a long enough supply chain (3 tiers was enough), and a long enough feedback loop (it turns out that even 1.5 hours was enough to demonstrate this), smart, intelligent people put into a structurally untenable situation would make a hash of things, even as they were doing their darnedest to optimize their individual roles and react to the environment. The minute you see this in action you will learn to stop blaming the individuals who are acting in their roles, and start analyzing structure and decision making in organizations from a feedback/information perspective.

The Beer Game also taught me to see how important open-ness and widespread dissemination of information was inside organizations, and I think Google in particular was very good at staying open despite all the pressures to do otherwise. Everyone I spoke to knew that we had a problem, and everyone was aware that the person responsible for trying to solve the problem was trying to do their best. This made organizational change much easier than it would have been otherwise (though it still wasn't easy). When confronted with situations like this, organizations tend to want to find an outside savior: someone who can ride in on a white horse from outside the company to save it. However, this is precisely the wrong approach, because the real problem usually has to do with the organization's approach to the problem, and that's much more easily changed by an insider.

By far the thing that impressed me the most was what a skilled organizational hacker Wayne Rosing was. When he saw that the previous approaches wasn't working, Rosing decided to take the problem and hand it to an engineer with previous experience in the domain! In other words, everyone knew that if Ron Dolin came and asked you for help, this was an important situation and critical for the IPO. If Wayne Rosing was at Google today, I'm convinced he would take that approach again with respect to a "Head of Social" position, which is to take an engineer well-versed in the social world and ask him to fix the problem. I don't know who that would be (my nomination would be Matt Cutts) but I'm positive it would be a better solution than a search for an outside Head of Social. The need in this case isn't for someone with external credibility, but for someone who can hack the organization from the inside to recognize that the important problems are the unknown unknowns.
Post a Comment