Thread by @crystalwidjaja

the data-informed product manager 101 toolkit:

1/ XmR Charts: identify whether changes in metrics are significant or just routine variation that your boss shouldn't panic over 🤷‍♀️🤷‍♀️
i use (open source 👀)xmrit.com

2/ Abandonment Funnels: visualize user dropoff rates based on their last action 📈
helps determine whether a new feature actually improved funnel conversion rates or you just moved dropoffs further down the funnel 🤷‍♀️ even better when segmented for side-by-side comparisons

3/ Altitude Maps: kpi trees on steroids 🌎
what are your manager's goals, your KPIs, problems you solve, and scope of your opportunities? size the potential of each problem to see if they're really an opportunity.

here's a step-by-step process:
1⃣ what % of your users dropoff at each step of your funnel?
2⃣ breakdown the dropoffs at each step into quantifiable segments (exit survey answers, error logs, last clicked feature, deviceTypes...)
3⃣ generate as many fan theories as possible by comparing the behaviors of successful vs unsuccessful users
4⃣validate at least one question: what needs to be true for this to be the biggest reason for the dropoff identified in 1⃣?

4/ Risk Ratios: using that Altitude Map, quantify each fan theory (a well-sized problem = an actual opportunity) and its relationship to your desired outcome by calculating their Risk Ratio. this achieves two things:
1⃣ Identifies which variable to optimize (e.g. messages sent)
2⃣ Identifies the specific behavior of that variable that your users need to perform for it to lead to your desired outcome (e.g. at least 10 messages sent in the first 7 days).