Who This Helps
This is for you, the Junior Analyst who just saw a number go red. Maybe it's a 12% drop in weekly active users. Maybe it's a conversion dip that appeared overnight. Your job is to figure out why and tell someone what to do about it. No pressure, right?
The Product Portfolio Strategy course teaches leaders how to size bets and sequence work. But before you can do that, you need to know which bet is actually broken. That's where you come in.
Mini Case
Imagine you're looking at your product's daily sign-ups. Last week they averaged 1,200. This week they're at 950. That's a 21% drop in 7 days. Your boss wants answers by Friday.
You check the data. The drop is concentrated in one region. Users from that region are hitting a new error screen after clicking "Sign Up." The error rate jumped from 2% to 18% in three days. You found the root cause: a recent deployment broke the sign-up flow for that region's language setting.
Now you have a clear recommendation: roll back the deployment or fix the language-specific bug. Your analysis is clean, your recommendation is actionable, and you saved the team from chasing ghosts.
Do This Now (5 Steps)
- Grab the last 30 days of data for your KPI. Look for the exact point where it dropped. Was it sudden or gradual?
- Segment by region, device, or user type. The drop might hide in one slice. In the case above, the drop only showed up in one region.
- Check for recent changes. Did a new feature ship? Did a deployment happen? Did a third-party API update? Match the timing.
- Talk to one person on the team. Ask them: "What changed right before the drop?" They might remember something the logs don't show.
- Write one recommendation. Keep it short. Example: "Roll back the sign-up flow deployment from Tuesday. This should restore conversion to baseline within 24 hours."
Avoid These Traps
- Don't blame the data first. A 12% drop might be a real problem, not a tracking bug. Investigate before you dismiss it.
- Don't jump to the biggest segment. The drop might be tiny in aggregate but huge in a small slice. Look at all slices.
- Don't write a novel. Your analysis should fit on one page. Your recommendation should fit in one sentence.
- Don't ignore the easy fix. Sometimes the answer is a configuration change, not a full product overhaul.
- Don't forget to check the date range. A 7-day drop might be a weekend effect. Compare same day of week.
- Don't assume correlation is causation. Just because sign-ups dropped after a marketing campaign doesn't mean the campaign caused it.
- Don't wait for perfect data. You have enough to start. Ship the analysis, then refine.
- Don't skip the recommendation. Your job isn't just to find the problem. It's to suggest what to do next.
Your Win by Friday
By Friday, you will have diagnosed the root cause of one KPI drop, written a one-page analysis, and delivered a clear recommendation to your team. You'll feel like a detective who just cracked the case. And hey, you might even get a high-five from your lead.
Remember: the Product Portfolio Strategy course helps leaders decide which bets to place. Your analysis gives them the confidence to place the right one. That's a win for everyone.