# Analytics tool recommendation: Acme Reports (fictional example)

Date: 2026-10-17. Prepared for: Jane Doe, founder (jane@example.com). Status: proposed.

Acme Reports is a fictional 12-person B2B SaaS company, two years old, seed-funded with USD 3M raised, selling a reporting product to teams. Every name and number here is invented.

## Answer

- Use: GA4 for site landing pages and the Google Ads `sign_up` key event (founder). One product tool on app and backend for workspace questions (PM): pilot Amplitude and Mixpanel on one workspace fixture.
- Do not buy: a second product tool, PostHog for now, Heap, replay, experiments, guides, warehouse-native analytics.
- Cost: USD 0 in year one on either startup programme, if accepted. Month 13 at growth volume: Amplitude Plus USD 0 if its group analysis passes, else a Growth quote; Mixpanel USD 78.40 a month. Checked on 2026-10-17.

## Decisions behind it

- Channels bringing trial workspaces that report within 7 days. Operator founder; unit workspace; gap workflow cost (weekly manual join).
- Where workspaces stall connecting a data source. Operator PM; unit workspace; gap data and capability.
- What broke after a release. Operator engineer; unit person and session; gap data: no release version on events, failed exports unrecorded.
- Fix first (tracking-plan-builder): backend `data_source_connected` and `export_failed` events; `workspace_id` and `app_version` on every event.

## Not buying, and why

- PostHog: dropped at step 3; the decisions are the PM's and the release question is a data gap. Growth USD 63.80, above Amplitude Plus at USD 0. If Amplitude fails the fixture, PostHog undercuts Mixpanel and gets the PM trial task first.
- Heap: fails step 4 (account analytics from Pro); 15,000 app sessions against a 10,000 Free allowance.
- Replay, experiments, guides, a second product tool: no owner or decision. Warehouse-native: no models.

## Billing meter and forecast

- Amplitude Plus (events): normal 400,000 and growth 1.2M, USD 0 (first 2M free). Fault 3.0M (autocapture): USD 94 monthly or USD 75 a month yearly, plus USD 28 if Accounts is needed and sold on Plus. Month 13: USD 0, or the Growth quote. MTU pricing costs more (appendix).
- Mixpanel (events at ingestion): normal USD 0; growth USD 56 overage plus USD 22.40 Group Analytics on the 1M plan; fault 3M tier USD 378 plus the quoted add-on; month 13 USD 78.40.
- PostHog (all app events identified): normal USD 0, growth USD 63.80, fault USD 453.

## Two tools: dual-tracking cost

GA4 carries site page views and `sign_up` only: 8 hours once, then 2 hours a month at a USD 90 planning rate (USD 180). Jobs only GA4 does: Ads bidding, remarketing audiences, landing-page reporting with the product SDK off the site.

One-tool option: Ads tag through GTM with enhanced conversions, Ads-tag audiences, one window as a secondary action: 16 hours once (USD 1,440), payback 8 months. Not taken: GA4 still owns site reporting, which would otherwise put 40,000 anonymous sessions a month on the product meter.

## Limits and alerts

- Amplitude Plus: overage may bill at 1.0x; suspension possible above 200% of purchase. Alert on monthly event volume at 1.6M (80% of 2M), owner PM.
- Mixpanel Growth: no cap; emails start at 85%. Usage report alert at 80% of forecast, owner PM.
- Flags: not in scope.

## Dependencies kept

- Google Ads `sign_up` conversion and remarketing audiences: GA4, owner founder.
- Weekly founder report: rebuilt on workspaces from first-touch signup properties.

## Next steps

- Tracking fixes: engineer, by 2026-10-31.
- Apply to both startup programmes; pilot both for 21 days, PM as operator. Test Amplitude on Plus-level group analysis, not Growth year-one access.
- Record in `martech-context/decisions.md` and `stack.md` after approval.

## Revisit when

- Usage passes 80% of forecast; a startup year ends (quote 60 days before); a second team needs governed definitions or experiment results; a tool goes two quarterly reviews uncited.

## Appendix A: forecast arithmetic

- Product events: 3,000 people x 5 sessions x 20 events + 100,000 backend = 400,000 a month. Growth (three times the people): 1.2 million. Fault: 9,000 x 5 x 60 + 300,000 = 3.0 million.
- Amplitude Plus estimator data, checked on 2026-10-17, amplitude.com/pricing: 2M events USD 0; 3M USD 94 monthly or USD 900 a year (USD 75 a month). The Plus data carries an Accounts add-on line, USD 0 at 2M and USD 28 a month at 3M, while the plan comparison lists Accounts for Growth and Enterprise; written confirmation requested.
- Amplitude MTU check, same source: growth has 9,000 identities at about 133 events each. The cheapest MTU purchase, 10,000 MTUs, is USD 160 against USD 0 on events. Normal: 2,500 MTUs at USD 84 bills 3,000 (120%), about USD 100.80, against USD 0.
- Mixpanel growth: 200,000 events over the 1M plan x USD 0.00028 = USD 56.00; Group Analytics at 40% of that overage = USD 22.40; total USD 78.40 a month. The 1.5M tier would cost USD 140 billed monthly or USD 100 a month billed yearly, plus its add-on, so the overage route wins. Annual billing needs more than 1M committed monthly events, so this route is monthly. Checked on 2026-10-17, docs.mixpanel.com/docs/pricing and mixpanel.com/pricing.
- Mixpanel fault: overage on the 1M plan would be 2M x 0.00028 = USD 560 plus USD 224 Group Analytics; the 3M tier (USD 378 billed monthly, USD 270 a month billed yearly) is cheaper, and the billing page allows upgrading before the cycle ends.
- Mixpanel normal: inside the free 1M; Group Analytics at 40% of zero overage. Confirm in the quote that no minimum applies.
- PostHog growth: 200,000 x 0.00005 + 200,000 x 0.000198 + 200,000 x 0.000071 = 10.00 + 39.60 + 14.20 = USD 63.80. Fault: 84.30 + 267.70 + 101.00 = USD 453.00; without group analytics USD 352.00. Checked on 2026-10-17, posthog.com/pricing.
- Like for like at 3.0M with account analysis: Amplitude Plus USD 94 plus up to USD 28; Mixpanel USD 378 plus the quoted tier add-on (the 40% rate applies to overage, not to pre-purchased tiers); PostHog USD 453.
- GA4 keep or replace: 2 hours x USD 90 = USD 180 a month to keep; 16 hours x USD 90 = USD 1,440 once to replace the Ads role; 1,440 / 180 = 8 months. Replacement passes the 12-month payback but fails the "no other job" condition, so GA4 stays.
- Heap: 3,000 x 5 = 15,000 app sessions; with the site 55,000. heap.io/pricing.

## Appendix B: hard-requirement evidence

- Workspace funnel with several members: Mixpanel Free fails; Growth with Group Analytics passes pending the fixture, and the Startup Plan includes it. Amplitude Free fails; Plus lists group behaviour analysis, pending the fixture and written confirmation of whether the Accounts package is needed. PostHog with group analytics passes pending the fixture. Heap Free and Growth fail. Checked on 2026-10-17: mixpanel.com/pricing, docs.mixpanel.com/docs/pricing/startup-program, amplitude.com/pricing, posthog.com/docs/product-analytics/group-analytics, heap.io/pricing.
- Saved behavioural cohorts for the weekly workspace report: Amplitude Plus 20 per organisation and Mixpanel Growth pass; Amplitude Free (none) and Mixpanel Free (cannot save) fail. Checked on 2026-10-17, amplitude.com/pricing, mixpanel.com/pricing.
- Google Ads conversions and audiences: kept in GA4. Checked on 2026-10-17, support.google.com/analytics/answer/9379420.
- At least 12 months of history for explorations: GA4 standard offers 2 or 14 months; set 14 months (a human admin change). A property reclassified as Large drops to 2 months and loses older event data. Amplitude Plus keeps 2 years of data access. Checked on 2026-10-17, support.google.com/analytics/answer/7667196, amplitude.com/pricing.
- Data region and regulated data: not required. Acme processes no health data and accepts US hosting.
- Startup programmes: Mixpanel (under 5 years, up to USD 8M raised, never on a paid plan) appears to apply; Amplitude (under 20 employees, under USD 10M) appears to apply; PostHog (under 2 years, under USD 5M) does not. Each vendor decides. Checked on 2026-10-17 on each programme page.