Skip to content

Turn Raw Weekly Metrics Tables Into a Clear 'What Moved, Why, and Next Action' Brief

Stop pasting silent Looker screenshots into Slack hoping leaders interpret them—explain true variance vs noise, isolate cohort drivers, and name the exact owner acting.

Copy the prompt and paste it into the AI assistant you already use.

What separates real analytical commentary from 'reading the chart aloud'

  1. 01

    Never write 'Demos fell 18% from 50 to 41' and stop—state the Driver

    Executives can read `41 vs 50` themselves; your entire value is decomposing *where* the -9 happened ('8 of the -9 dip came from EU Paid Search after Wednesday's landing page test; US Organic held flat').

  2. 02

    Separate 'Structural Trend' from 'Calendar / Whale Noise'

    Always normalize for fewer business days (Thanksgiving/bank holidays), end-of-month billing clumps, or a single $50k M&A churn event so teams don't panic-pivot on statistical noise.

  3. 03

    Compare against BOTH WoW/MoM AND Plan/Pacing Target

    Being `+12% WoW` after a terrible prior week can still leave you `-15% behind Q4 pacing`—always pair short-term delta with quarter-to-date attainment.

  4. 04

    End every negative variance bullet with `-> Action + Owner + Check Date`

    Close every red metric with what is already happening: `-> Action: Growth reverted EU LP variant B Friday night (@Leo); verifying rebound by Wed 12 PM.`

How it works in practice

Raw BI dashboard numbers pasted in Slack

Week 45: Trials = 310 (-4% WoW). Day-7 Activation = 24.2% (down from 31.0% W44; Q4 target 30%). Paid New ARR = $64k (+8% vs plan).

Executive Weekly KPI Narrative

HEADLINE: Paid ARR paced +8% above plan ($64k), while a Thursday OAuth bug depressed Day-7 Activation (-6.8 pts)—fix deployed. • 🔴 DAY-7 ACTIVATION: 24.2% (-6.8 pts WoW | Target: 30%) - Driver: Microsoft 365 signup flow threw scope error Thu–Fri (0% activation on 68 M365 signups; Google Workspace held steady at 31.4%). - Action: Hotfix shipped Sat 09:00 UTC (@Dan); Support emailed 68 affected users magic retry link Mon AM. • 🟢 NEW PAID ARR: $64k (+8% vs weekly pacing; QTD at 103% of plan). • ⚪ TRIAL VOLUME: 310 (-4% WoW — normal variance post-conference spike).

Splitting Microsoft 365 vs Google Workspace proves core onboarding isn't broken—preventing a pointless product redesign debate.

Common questions

What if a metric moved sharply this week and I don't know the root cause yet?

Never invent a plausible guess—write explicitly: `Driver under investigation: ruled out tracking pixel & geo mix; auditing pricing-page cohort split today (@Maya, update Tue 2 PM).`

Should numbers sit inside paragraphs or in a compact summary table?

Put exact numbers in a clean 5-row scannable table (`Metric | Actual | vs Plan | WoW`) and reserve bullets strictly for 3 high-signal variance stories.

How many metrics belong in a weekly executive digest?

Five to eight North Star and primary funnel indicators max; link your 40-chart operational dashboard for deep exploration.

Learn more

Related guides and background articles for this task