Who This Helps
This is for junior analysts who get a Slack ping saying "revenue dropped 12% this week" and need to figure out why—fast. You want to ship a clean analysis with clear recommendations, not a 20-slide deck full of noise.
Mini Case
Meet Priya, a junior analyst at a SaaS company. Monday morning, her VP asks: "Why did our trial-to-paid conversion drop from 22% to 18% in the last 7 days?" Priya uses the approach below. In one focused session, she finds the root cause: a pricing page change that confused mobile users. She ships a one-pager with a fix recommendation. Her VP approves the change by Wednesday.
Do This Now (5 Steps)
- Define the drop window and baseline. Pick the exact time period (e.g., last 7 days) and compare it to the prior 7 days. Use a simple table: metric, before, after, change.
- Segment by user type. Break the metric by channel, device, or plan. In Priya's case, mobile users dropped 30% while desktop stayed flat. That's your first clue.
- Look for event-level changes. Check if a specific step in the funnel changed. Did fewer users reach the pricing page? Or did they reach it but not convert? Priya saw the pricing page views were normal, but conversion from that page fell.
- Check recent releases or changes. Ask: Did we push a code change, update copy, or run a new campaign? Priya found a pricing page redesign went live 8 days ago. Bingo.
- Write a one-page diagnosis. State the problem, the root cause, the evidence (with numbers), and one clear recommendation. Keep it to 5 bullet points max. Your VP will love you.
Avoid These Traps
- Don't analyze everything. Focus on the one metric that dropped. Ignore vanity metrics.
- Don't blame a single factor without evidence. Correlation is not causation. Check multiple segments.
- Don't skip the "what changed" step. Most root causes are recent changes. Ask around.
- Don't write a novel. One page, clear actions. No one reads long reports.
- Don't forget the recommendation. Diagnosis without a fix is just complaining.
Your Win by Friday
By Friday, you'll have shipped a clean analysis that pinpoints the root cause of a KPI drop. Your team will know exactly what to fix. You'll look like the analyst who gets things done. And you'll have saved yourself from a week of spreadsheet hell. That's a win.