Diagnose a KPI Drop Like a Junior Analyst

Who This Helps

This is for junior analysts who get a Slack message saying "revenue dropped 12% this week." Your boss wants answers by Friday. No pressure. You need a repeatable way to find the real problem and suggest what to do next. The Product Portfolio Strategy course teaches you to size bets and sequence work, so you can turn a scary number into a clear story.

Mini Case

Imagine you work at a subscription app. Last week, new sign-ups dropped 12%. Your first instinct is to blame the marketing team. But after digging, you find the real issue: a pricing page bug that broke the checkout flow for 3 days. The fix took 2 hours. The lesson? Don't guess. Diagnose first.

Do This Now (5 Steps)

1. Get the raw data. Pull the KPI for the last 30 days. Look for the exact day the drop started. That's your clue.
2. Segment the drop. Break it by channel, device, or region. In our case, the drop was 80% on mobile Safari. Now you know where to look.
3. Check recent changes. Did the team deploy something? Change a price? Update a page? Ask in Slack. The bug was introduced 3 days before the drop.
4. Run a quick correlation. Compare the KPI with other metrics like page load time or error rate. The error rate spiked 200% on the same day. Bingo.
5. Write one recommendation. Don't list 10 fixes. Pick the one that matters most. "Fix the pricing page checkout button on mobile Safari." That's a clean analysis.

Avoid These Traps

• Blame the first thing you see. Marketing might be innocent. Look at data first.
• Skip the timeline. Without knowing when the drop started, you're guessing.
• Ignore small segments. A 12% drop might hide a 50% drop in one city.
• Write a novel. Your boss wants a 3-sentence summary, not a 10-page report.
• Forget the fix. Diagnosis without a recommendation is just complaining.
• Assume it's a one-time thing. Check if the pattern repeats weekly.
• Don't ask for help. A teammate might have seen the bug already.
• Overthink it. Most drops have a simple cause. Start there.

Your Win by Friday

By Friday, you'll have a one-page analysis that says: "The 12% drop in sign-ups was caused by a checkout bug on mobile Safari. Fix it in 2 hours. Revenue should recover in 3 days." Your boss will nod. You'll feel like a detective. And you'll have a repeatable process for next time. That's the power of diagnosing a KPI drop the right way.