As a Product Manager data is your best friend and you should treat it like a best friend, well kind of… there’s no need to take it out for beers and see how it’s doing, but you should see how it’s getting on.
Data in all its forms will give you information, evidence, ideas, shocks, gold and restful nights. Make sure your data is accessible, make sure it’s clean and make sure you connect data sources together or if you really can’t connect them, then marry two sources of exported data together manually.
Here are my top 10 tips for data.
1. Get into what data you have, get dirty with it, find it’s problems, issues and crazy set up. We’ve all seen bad data, you need to know how good or bad yours is.
2. Add to the data you have, fill the holes in it as early as you can, if it’s missing a created date, go get one added ASAP.
3. Missing a type of data? Get it added. A great example here is front end event reporting, my last role didn’t have any reporting here, so I got Amplitude installed as quickly as I could.
4. Save everything. It’s never too early or too late to save data, go get a Mongo database or some awesome Amazon tool and save save save, data is an asset.
5. Dig deep into big data (your saved data above), it will yield gold, I promise.
6. Plan reporting of a feature when you build a feature. If you think you’ll need to know when a user last used a feature, then add LastAccessedDate to your specifications.
7. Document how you get to a metric. Let’s say you report on ‘Users with green hair’ make sure everyone has a single point of truth to how they also create that cohort, so their reporting matches your reporting.
8. Dedicate time to speak with real users, document their answers in a controlled centralised manner, so future PMs (and you) can use this too.
9. Bringing in specialists, digging deep in data is a skill and it needs time, if you don’t have the time, then bring someone in that does. It will pay for itself.
10. Document your experiments, the hypothesis and their results, somewhere central, so when Dave says “We should try this” you can reference the previous experiments as an example to not do it, or to help you look at it a different way.
If you do all of this with your data it will surely become your best friend, so much so that you’ll have a little time to take a real friend out for those beers, now doesn’t that sound like a good use of data?!
- I started this blog just to share my workplace opinions and hope to help some people new or old in the industryI have worked in IT, eCommerce, Corporate digital, a start ups for over 20 years now and have an abundance of experience from many sectors.I have been lucky to…