I recently was speaking with my dad about his friend’s business. This friend and my dad go way back, having successfully started a number of businesses over the years. They’re now doing a victory lap, and it’s so fun to see my dad as excited in retirement as I imagine he was at my age, but I digress. We were talking about how they wanted to start offering performance incentives, and I had a few things to say about the topic. I’ve slightly edited an email I sent him and offered it here for all.
—
You mentioned performance incentives & gave a quick example of what that might look like, e.g. what behavior you’d incentivize. The example was “time to clean a room”, which if I’m not mistaken, had some weird name at your business, but I’ll call it… cleaning a room. If you monitor "cleaning a room” time then you’re very explicitly creating a rule that everyone is incentivized to play by.
Any time you have a rule, you have a game.
And what do all of us do with rules in games? We stretch them, or even cheat. Everyone you’ve ever thought of as a good person has cheated on something they thought didn’t matter, or at least, didn’t matter to them, especially if they didn’t think they’d get caught. I can imagine anyone starting to cut corners in the clean and quality starting to dip. That to me seemed off-brand for your business, which is basically about quality above all else. Maybe it doesn’t matter how quickly you clean a room and it matters more how clean it is. You might be better off measuring that.
Peter Drucker has a lot of phrases I’m sure we’ve both heard and used, one of my favorites is “what get’s measured gets managed”. The typical interpretation is that if you want to manage something, start measuring it. I take it as more of a warning. Whatever you measure had damn sure better be the thing you want to manage.
If that was the metric, here’s how I’d approach creating the incentive. Sidenote: there’s also pretty good evidence that incentives don’t really work, but if you’re gonna do them, I think incentivizing the right thing takes a thoughtful approach).
First, involve the people doing the work in setting the standards and procedures, and give them ownership in the process. No one likes following rules that they don't understand, but if you allow them to come up with the rules you’ll probably be delightfully surprised by how much better they are and how much more excited they are to follow them. This is another tried and true concept I stole from somewhere (I think a marine), but the gist is that the person closest to the problem should be trusted to know best how it can and should be solved.
Next, give them a challenge focused on quality, not time. In this case, you could ask them to write a checklist of the things you do when cleaning a room that leads to a quality cleaning, then ask them to prioritize that list. Have them use it a few times and ask them to make changes.
Then start using it for every cleaning and track how long it takes to go through the checklist. I’m guessing you’d measure a dramatic decrease in time (increase in efficiency) over the first few, and then see a leveling out. I’d use that as a standard bar for anyone who’s done a decent number of cleanings, and use it to see how quickly newbies adapt or if someone is slipping. I’d be careful about trying to continuously squeak out second after second unless you’re using new technology or reducing the scope of the work. If a crew is regularly within a small margin of their target, reward them.
—
To summarize, I’d be very careful what you incentivize, and if you choose to incentivize certain behavior, reward consistency not one-time performance. I guess I could’ve just said that, but I think the context was important. In fact, I had to try really hard to not go down a number of additional tangents. But wonder if you, the reader, would be curious to read any of those tangents. I’ll make a brief list below - please let me know which, if any, you might like to hear more about. I love writing but I need someone to write to in order to make it good. I’ll phrase these as questions because that’s kind of how they appeared in my head as I was writing.
If incentives don’t work, what should I do instead?
How do you choose what to measure to get the result you want?
When is measuring time bad? Why do we default to measuring time reduction as “progress”?
If creating rules makes a game, how do we design good games?
How much autonomy is too much? What factors into that determination?
If I don’t hear anything, I’ll probably just pick one to write about next. How’s that for an incentive?