Growth Engineering: Events, Identity, and Experiments

Part I — Foundations

1. 1Tracking is for optimisers, not dashboardsWhat growth engineering is and is not.
2. 2The LoopGrowth models, loops versus funnels, AARRR, and the growth accounting equation.
3. 3The MetricNorth-star metrics, input metrics, guardrails, and why vanity metrics survive.

Part II — Instrumentation

1. 4The EventEvent design, the tracking plan as a contract, and what to track.
2. 5Event semanticsWhat an event asserts, when it fires, and what an optimiser learns from it.
3. 6The IdentityAnonymous IDs, aliasing, merge, and the user, account, and workspace problem.
4. 7Context and traitsAttribution source, experiment variant, and plan: how outcomes become attributable to causes.
5. 8The PipelineSegment, RudderStack, and the collect-once, route-everywhere model.
6. 9Client-side versus server-side trackingAd blockers, consent, and the data loss you do not know about.
7. 10The SchemaProtocols, versioning, governance, and who owns a broken event.

Part III — The Data Stack

1. 11Warehouses, ELT, and martsWhere definitions get frozen into tables.
2. 12The DestinationWhat each consumer needs, and the contracts they impose.
3. 13Reverse ETL and activationHow warehouse data gets back into the tools.
4. 14Product analytics toolsAmplitude, Mixpanel, PostHog, and when to skip them.
5. 15AttributionUTMs, click IDs, first-touch, last-touch, multi-touch, and the limits of honesty.

Part IV — Experimentation

1. 16The ExperimentAssignment, exposure logging, feature flags, and cohort semantics for tests.
2. 17Statistics for engineersPower, minimum detectable effect, peeking, and sequential tests.
3. 18Metric pipelines and sample ratio mismatchSample ratio mismatch and the other ways your test is lying.
4. 19Experiment platformsBuild versus buy: Optimizely, Statsig, LaunchDarkly, and in-house.
5. 20When not to testSmall samples, long-cycle B2B, and quasi-experiments.

Part V — Growth Surfaces

1. 21The TriggerRules that turn events and audiences into actions.
2. 22Onboarding and activation engineeringThe first session as an engineering problem.
3. 23Lifecycle messagingEmail, push, in-app, and the orchestration layer: Braze and Customer.io.
4. 24Ad platforms as optimisersConversion APIs, deduplication, and how to feed the algorithm the right signal.
5. 25Referral, virality, and PLG mechanicsThe loops that run on the product itself.
6. 26Pricing, paywalls, and monetisation experimentsWhere the money meets the metric.
7. 27SEO and programmatic pages as engineering problemsSearch as a growth surface you can instrument.

Part VI — Behavioural Systems

1. 28The AudienceCohort semantics, bounded versus unbounded, refresh, and what "active" means.
2. 29Segmentation and user classificationRules, scoring, ML, and how to keep it explainable.
3. 30The CRM as a destinationHubSpot, PQLs, and how to hand signals to sales.
4. 31Personalisation and the recommendation trapWhen adaptive surfaces help and when they hide the signal.

Part VII — Operating the Function

1. 32Org designEmbedded versus centralised growth teams.
2. 33VelocityThe experiment backlog, prioritisation, and why ICE and RICE are fine.
3. 34Privacy, consent, and retentionGDPR, CCPA, consent, and data retention as first-class engineering constraints.
4. 35Technical debt in growth codeSunsetting experiments, flag hygiene, and dead events.
5. 36EthicsDark patterns and the long-term cost of short-term wins.

Appendices

1. AA reference tracking planA complete tracking plan for a fictional product, with every event and property justified.
2. BExperiment readout templateThe one-page readout that a test must produce before anyone acts on it.
3. CGlossaryEvery term in the book, defined once.
4. DThe primitivesEvent, Identity, Context, Schema, Pipeline, Destination, Metric, Experiment, Trigger, Audience, Loop. Built from PRIMITIVES.md.