What should be done?
Step 1 is to fix your hiring process, which is probably a low-context slopfest. Ever heard the expression “Garbage in, garbage out?” That’s the inherent issue of most businesses, in all honesty. You recruit like shit and have ridiculous expectations but low salaries? Well, you get shit. And when you get shit, your business won’t be that productive. I don’t care if you have the best product in the world. If morons are in every crevice of your org, eventually it collapses. So fix your recruiting process. That’s No. 1. Here’s one idea on that.
Step 2 is that the internal culture cannot be all about hierarchy and “get your work done” and “I’m so busy.” There has to be room to ask questions, push back, learn about other teams/departments, etc. Without that it’s all silo garbage. That means the data collected will just be silo’ed to the hilt and the end experience for whoever uses your stuff will be a mess. No bueno.
Step 3 is to collect and scrub the data. Notice you don’t do this until your hiring and culture are humming.
Step 4 is to figure out how to present the data and what it means to the people who can make decisions. What formats? How should the presentations be? Long/short? You absolutely need to think about this. I had a job once where some dude presented to the senior leadership with a great data deck. He had worked so hard on it! They didn’t understand 3/4 of it and all his work essentially went nowhere. These are guys who want words they understand (“CAGR”) in formats they like (PPT). Don’t fuck with that. It’ll be too jarring for them. Your hard work goes nowhere.
(Related to all this: what if we could turn truck drivers who get automated out into data scientists? We might be able to.)
Step 5 is to make a few big, impactful, efficient decisions based on analytics. Then spread the gospel of how it happened. Talk it up around the company. Do some knowledge sharing around it. Make people believe in the whole deal. The problem with data/analytics at many companies is that it seems clustered to one department and maybe the execs. It’s not “fully owned.” To make it actually work for the org, it needs to be owned by multiple people — not just some dude from Stanford and his direct reports.
Step 6 is making money. I know you had to wait six steps, and that’s a bitch. But if you fuck up the first five steps, this one will be a lot harder.
What else you got / how could it be better?