The Analytics Trap
You're three weeks into your startup. Someone says: "We need Amplitude, Mixpanel, Tableau, and custom Looker dashboards." Suddenly you're paying $2,000/month for tools. Your data engineer spends 40 hours setting up funnels. No one looks at them.
This is the standard failure mode. Scott Belsky calls it "analysis paralysis in the Messy Middle." Too many dashboards, too little conviction.
YC companies do the opposite. They obsess over five metrics on a single sheet. They review weekly. They move fast because they're not drowning.
The One-Dashboard Rule
One dashboard. Not one per department. One.
Use Google Sheets. Use Excel. Use Metabase if you must. The tool doesn't matter. The discipline does.
Your dashboard lives here:
- Public to the whole team.
- Updated every Monday morning.
- One tab. One view.
If you can't print it and understand it in 60 seconds, it's too complex.
The Five Metrics Framework (AARRR)
YC's standard: AARRR (Acquisition, Activation, Retention, Revenue, Referral). Pick one metric per layer. Not three. One.
Acquisition: Monthly Active Users or signups.
- For B2B SaaS in India: Track by geography (Bangalore, Mumbai, Hyderabad matter differently).
- Typical benchmark: 10-30% MoM growth at Series A stage.
Activation: % of signups completing core action in first week.
- Core action = first project created, first export run, first payment method added.
- For Indian SaaS: 25-40% activation is solid (India's mobile-first users drop faster).
Retention: % of users active after 30 days.
- Most critical. YC data shows retention at Day 30 predicts revenue better than signup velocity.
- Indian B2B: 60%+ is good. Indian B2C: 20%+ is survival.
Revenue: MRR (Monthly Recurring Revenue) or ARR.
- If pre-launch: Track contract value instead (even if unsigned).
- For India early-stage: Show both USD equivalent and INR to match investor expectations.
Referral: % of signups from existing users or word-of-mouth.
- Most startups ignore this. It predicts product-market fit.
- If this is 0%, your product solves a problem only you care about.
Why five? Not three, not ten.
Three metrics hide problems. Ten metrics hide decisions. Five forces prioritization.
The Weekly Review Ritual
Every Monday, 10 AM. Non-negotiable.
The 15-minute meeting:
1. Read last week's narrative (you wrote this last week).
2. Update the five metrics.
3. Answer: What surprised us?
4. Write next week's narrative in two sentences.
5. Assign one action if something broke.
That's it. No PowerPoint. No deep dives unless a metric moved 20%+.
Why weekly? Longer intervals hide inflection points. Shorter intervals create noise. Weekly data compounds visibility—52 data points per year reveals seasonality, viral moments, and cohort quality.
What NOT to Track
- Vanity metrics (total signups ever, page views).
- Metrics you can't act on ("average session duration" without knowing why it changed).
- Proxy metrics masquerading as real metrics ("engagement score" instead of "time to first revenue").
- Competitor metrics (let them distract themselves).
Indian founders often track this: "How many impressions did our Product Hunt launch get?" (Vanity.) Track instead: "What % became paying users?" (Actionable.)
The Benchmark Reality
Your metrics are worthless without baselines.
SaaS (B2B), India, Pre-PMF:
- Activation: 20-35%
- CAC: ₹3,000-₹15,000 (depends on sales vs self-serve)
- CAC payback: 4-6 months (Indian SaaS recovers slower than US)
- Retention (Day 30): 40-60%
SaaS (B2B), India, Post-PMF:
- Activation: 50%+
- CAC payback: 2-3 months
- Retention (Day 30): 70%+
- MRR growth: 10%+ monthly
If you're 2x worse than post-PMF benchmarks, your problem isn't metrics. It's product.
Non-Obvious Insight: The Cohort Trap
Most founders track aggregate metrics ("50% of users retained"). Better: track by cohort.
Your January users might have 70% Day 30 retention. February users: 45%. Why? Your onboarding changed, or your product shifted, or January got a better feature. You only see this if you separate cohorts.
Google Sheets cohort table: 30 minutes to build. Reveals everything.
The Setup (Actual Steps)
1. Pick your five metrics.
- Use the AARRR framework above. Don't customize yet.
2. Where does data live?
- Signup/activation: Your product DB or Amplitude free tier.
- Retention: Same.
- Revenue: Stripe API or your billing system.
- Referral: Check user's `source` field at signup.
3. Create the sheet (2 hours).
- Rows: Dates (by week or month).
- Columns: Your five metrics + notes column.
- Formula = pull from API or manually input (yes, manual is fine early-stage).
4. Automate later (not first).
- Week 1-4: Manual updates. You learn the data this way.
- Week 5: Use Zapier or custom script if updating takes >30 min.
- Never: Build a tool before you understand the metric.
Actionable Takeaway
This week: Create one Google Sheet. Add your five metrics. Add data for the last 8 weeks. Stare at it. What pattern do you see? That pattern is your company's heartbeat. Obsess over it. Everything else is noise.