At work things are good. I have a new job, which isn’t all that different from my last one. I enjoy coming to understand how work gets done though. Its a weird mix when people come together to do something and then a hierarchy emerges for who asks what to whom to get things done. I have learned a lesson I think about work. It might just apply to the technology development sector, but I suspect it applies to all large projects of a technical nature: coordination costs are HUGE.You have no idea the amount of effort that is spent trying to inform all relevant parties of stuff that other people already know…but you can’t ask those that already know because their calendars are already full the entire day. In other words, the cost of coordination have grown to the point that some people don’t do anything at all anymore…they just know general info that other people don’t have the time to know.

The value chain runs in a trickle from these people. They know what needs to be done…but because they’re in meetings all day, they can’t do anything really. They just trickle bits of relevant information to those below them, who further elaborate in trickles to those below them, until we actually get a usable output at the very bottom.

But those at the bottom aren’t even close to being equipped with the information to do it all on their own. It requires the small inputs of those above, in their small areas of expertise, to get to the bottom. The rub is that those above are only adding small values at each point in the chain. Each value add is small; which means each person at the top of the chain isn’t doing all that much, but their decisions require more and more information to make, which require more coordination costs…more meetings….more nothings.

All these people could be replaced by maybe two or three people that REALLY know how to do their work. This means people that have done perhaps all the jobs in the chain over a multiple year period. The problem is that these people don’t exist. No company can keep people working in a particular business long enough to staff the talent they need.

I always say, “Shit…I’ll take you, me and Johnny Smith…and we can do the work of all 15 of those clowns, and still go home at 3 o’clock everyday…and do it better.” And that’s true. But I could never get those three people to all work with me for any significant amount of time, because people that really know what they’re doing are in such demand that you can’t have a significant number of them working on the same project.

The staffing will always be terrible unless you find a way to pay (in some form, even if its not in $) these uber-producers. I don’t see any companies that are able to do that because its so hard to tease out who is actually driving the project. People know informally, but I’ve never seen a company that pays for it.

Imagine a low level manager where the pay is generally around $40,000…except some people, in the same job, are making $70,000…almost double. It’d never work. The other managers would get upset if they found out others were making so much. You’d need an airtight measurement system for incentives that everyone believed in. Most work is so fluid there is no way to measure that extra productivity…at least currently.

Here is a good idea for Google: they should try to figure out how to measure it.

Current metrics suck for large scale project related work. Its so hard to tease out who is driving results. Google had the same problem with the Internet: its so hard to tease out what someone means when they search for something. They figured out a better way.

They should figure out a better way to search for “who is driving results on this project”. Maybe it could be based on some weighted average formula of “number or emails sent to and from, number of meetings called/attended, number of rows of code written, number of phone calls to and from, number of mentions in other people’s emails, number of project tasks assigned, complexity of work (a tough one), # of positive mentions in the performance reviews, etc, etc”. Of course, the formula would have to be a secret or people would try to manipulate it (which happens with the current Google search). Regardless, its possible.

On all other fronts I am just chilling. I have posts coming on CYA, women and the workplace, and the placebo effect.

Leave a Reply