I'm sure everyone will tell you about Goodhart's law (anything you set as a target ceases to be a good measure because people will game it). The question is: w

I'm sure everyone will tell you about Goodhart's law (anything you set as a target ceases to be a good measure because people will game it). The question is: what is productivity and why do you want to measure it. Because the crux of the issue is whether what you're measuring really is productivity. The things that you are capable of measuring are often easy to measure, but poor indicators of productivity. And even then, is productivity what you want? The productivity needs to be aligned to the business goals. If you set engineers productivity targets they'll find the easiest way to meet those targets but they might not align to the business. But I think it's more important to look at how you actually manage people. The best way to get good work out of people is to make sure they have the information and resources they need to do the job, they're free from distractions, they have autonomy and authority to do what they need to get the job done, they're working in a high trust environment, and their achievement is recognised. None of those things are improved by measuring productivity, and trust, autonomy and recognition are actively hurt. By defining productivity for an engineer what you're saying is "You no longer decide how to best serve the company, I decide that, and you just follow what I say". Good luck, suddenly all the responsibility is on you - and you don't even control the outcome.