Dollar Driven

Company leadership say they're data driven, and they may even believe it. But the only data that actually matters is money.

Every other data you collect, visualize, and optimize is a proxy for money. If you're lucky, the data will be a direct translation to money, but most likely you'll have to go through two or more exchanges (managers) before you can get the real value. Any data that's worth collecting will have someone who translates it.

Customer sign ups, retention, feature usage, page views. It's all money.

Somewhere in marketing they're collecting social media impressions and trying to exchange that for revenue. Someone in ops is looking at log verbosity and trying to cut costs on their next s3 bill.

The best description I've heard of what a product manager does is they turn "data into dollars." The difference between a PM and most engineers is leadership will actually tell PMs how much money they make for the business.

Engineers are usually too many exchange layers away from money that leadership doesn't think it's worth telling them—or they can't be trusted. So they make up their own proxy data which usually starts with the easy things to collect (e.g. uptime, development velocity).

If you truly are data driven, and you're part of leadership, then you need to pass down the financial Rosetta Stone and share your financial details. The biggest surprise moving from Disney to Amazon was Amazon told me [some of] our financial data.

This not only help me make better decisions, it helped me realize what data was meaningless to collect. It also helped me realize there was nothing I could do at my level of responsibility to change the number in any meaningful way. The numbers were so big it just didn't matter.

I know they didn't openly share all of the financial information, but coming from Disney where we had to read about financial speculation from industry websites, it was refreshing to be trusted with some of the "real" data.

So no matter what data your collecting or what initiatives it's driving, try to translate it into the thing the data hoarders really care about. The next time you make a dashboard, think about how it'll get translated up the chain and you'll do a lot better.