We should have named a backup approver for production deploys before vacation season. In engineering pitfalls (thread 18): The smallest improvement to onboarding docs reduced repeated Slack questions. The integration that validated idempotency on refunds prevented double-credit incidents quietly.

The mentor who said 'tell me the risk in one sentence' sharpened planning instantly. In engineering pitfalls (thread 18): We should have deleted unused DNS CNAME chains pointing at deprecated marketing pages — drift hurts SEO too. We underestimated how long humans take to trust a new workflow.

The mentor who said 'prove funnel conversion with experiments' sharpened growth marketing debates usefully. In engineering pitfalls (thread 18): We stopped confusing 'busy calendars' with 'alignment' — fewer meetings, clearer notes won. Good incident comms reduce duplicate tickets more than faster fixes sometimes.

We finally admitted our test data did not represent production shape at all. In engineering pitfalls (thread 18): What saved us was a boring checklist, not another brainstorming session. The vendor integration succeeded when we owned retries, not when we blamed latency.

Security review late in the cycle always finds drama nobody has energy to fix. In engineering pitfalls (thread 18): The quiet win was documenting which API powers circle search vs user search — fewer wrong-team bug assignments weekly always. The mentor who said 'document the sharp edge' saved the next hire a week.

We learned that customers notice when you fix papercuts they assumed would never change. In engineering pitfalls (thread 18): Good telemetry feels like magic once you stop flying blind during incidents. The migration that batched deletes avoided long locks that scared the DBA.

We learned that 'no' to one thing is 'yes' to focus if you explain the trade. In engineering pitfalls (thread 18): We learned that naming owners for analytics pipelines prevents mysterious metric drift nobody owns. We stopped shipping 'just internal' spreadsheets as databases — they always become databases anyway.

The smallest permission boundary prevented a contractor from seeing the wrong dataset. In engineering pitfalls (thread 18): The architecture spike that listed rate limit strategy early prevented abusive traffic surprises in launch week. We learned that sustainable on-call means fixing root causes, not hero badges.

We learned that writing 'rollback criteria' in migration plans reduces bridge thrash at night. In engineering pitfalls (thread 18): We stopped shipping 'temporary' feature flags without removal tickets linked in Jira. We learned that naming runbook steps after people breeds single points of failure.

The mentor who said 'show me the user pain' ended a bikeshedding architecture thread. In engineering pitfalls (thread 18): We learned that repeating the same retro topics means we are not learning. Cut scope early; nobody remembers the sixth nice-to-have you skipped.

We should have deleted the feature nobody used; it still cost support time. In engineering pitfalls (thread 18): The integration that validated schema versions prevented silent consumer crashes on deploy. We stopped shipping 'temporary' reputation boosts for demos — they poison trust when members compare notes later honestly.

The design critique that asked about permissions on shared links prevented a leak. In engineering pitfalls (thread 18): The spreadsheet model of headcount lagged reality during hiring freezes. We stopped confusing 'community growth' with 'raw signups' when measuring circle health honestly.

The mentor who said 'tell me the rollback plan in one paragraph' improved migration quality. In engineering pitfalls (thread 18): We stopped confusing 'alignment meetings' with 'decision meetings' — different agendas, different outcomes. The architecture spike that time-boxed exploration prevented endless research tickets.

We optimised for demo day metrics and regretted it during the first real spike. In engineering pitfalls (thread 18): The flaky canary that ignored DB deadlock metrics taught us to watch locks not only HTTP 500s during migrations quietly measurably always. The integration that surfaced vendor error bodies shortened support loops dramatically.

We learned that humour in incidents is fine after service is stable, not before. In engineering pitfalls (thread 18): We learned that transparent pricing for paid circles beats hidden surcharges when hosts compare platforms quarterly helpfully finally always. The best postmortems include customer communication review, not only root cause.