tezvyn:

Growth & Experimentation

A/B testing, growth loops, conversion, retention

269 bites

Growth & Experimentation84 sec read

Set up a client-side button color A/B test

WHAT IT TESTS: practical experiment wiring on the client. OUTLINE: stable hashing of a persistent ID into buckets, conditional rendering of the variant, exposure plus click event logging. RED FLAG: re-randomizing on each load so users flicker between variants.

Growth & Experimentation2 min read

Attribute a mobile install to a desktop ad

WHAT IT TESTS: cross-device attribution methods and their trade-offs. OUTLINE: deterministic matching via a shared login is accurate but needs auth on both ends; probabilistic fingerprinting scales without login but is noisy and privacy-fraught.

Growth & Experimentation2 min read

Handle interaction effects on a shared page

WHAT IT TESTS: reasoning about interaction effects and mitigation. OUTLINE: combined variants may produce effects neither has alone; use mutual exclusion for likely interactions, orthogonal designs with interaction monitoring otherwise.

Growth & Experimentation2 min read

Why repeatedly extending a test inflates false positives

WHAT IT TESTS: understanding the peeking problem and inflated false positives. OUTLINE: repeatedly checking and extending until significance is p-hacking via optional stopping, which inflates the false-positive rate; fix with fixed sample sizes or sequential…

Growth & Experimentation2 min read

Architect a configurable, goal-based onboarding flow

WHAT IT TESTS: building a data-driven, configurable flow rather than hardcoded branches. OUTLINE: capture the goal, let the backend return a server-driven flow definition mapping goal to steps and content, render generic components on the client.

Growth & Experimentation2 min read

Increase experiment velocity for non-engineers

WHAT IT TESTS: decoupling experiment changes from release cycles safely. OUTLINE: server-driven config, feature flags, and a self-serve UI let non-engineers ship copy or layout variants instantly; add guardrails and metric checks.

Growth & Experimentation2 min read

Prevent conflicting experiments with layers

WHAT IT TESTS: enforcing mutual exclusion via layered assignment. OUTLINE: group conflicting experiments into one layer so a user's per-layer bucket maps to at most one of them; orthogonal layers can overlap.

Growth & Experimentation2 min read

Design a contamination-safe pricing experiment

WHAT IT TESTS: choosing a randomization unit that hides price differences. OUTLINE: user-level price tests leak via fairness perception, so use geo holdouts or time-based cohorts where everyone in a unit sees one price.

Growth & Experimentation2 min read

Run concurrent experiments without interference

WHAT IT TESTS: designing for parallel experiments via layered infrastructure. OUTLINE: independent non-interacting tests can share traffic through orthogonal layers; interacting ones need mutual exclusion in a shared layer.

Growth & Experimentation2 min read

Design a referral feature's lifecycle and races

WHAT IT TESTS: modeling a stateful flow with idempotency and concurrency safety. OUTLINE: a referral entity with explicit states, a unique constraint on the invited user, and atomic transactions plus idempotency to prevent double credits.

Growth & Experimentation2 min read

Why consistent user IDs matter in experiments

WHAT IT TESTS: understanding identity for stable assignment and clean measurement. OUTLINE: the user id seeds deterministic bucketing and ties events to one person across devices; achieve it via authenticated ids and anonymous-to-known stitching.

Growth & Experimentation2 min read

Trace an event from click to analysis

WHAT IT TESTS: end-to-end understanding of an analytics event pipeline. OUTLINE: client SDK captures and batches, a collection endpoint ingests, a stream and ETL enrich and load into a warehouse for analysis.

Growth & Experimentation2 min read

Communicate forecast uncertainty with prediction intervals

WHAT IT TESTS: quantifying and communicating forecast uncertainty. OUTLINE: a point estimate hides risk; produce a prediction interval via model error, simulation, or scenarios, and state assumptions.

Growth & Experimentation2 min read

When user-level A/B tests get contaminated

WHAT IT TESTS: recognizing interference that breaks the independence assumption. OUTLINE: network or marketplace spillover violates SUTVA, so randomize by cluster (geo, group, time) and analyze at that level.

Growth & Experimentation2 min read

Build an opportunity-sizing model before building

WHAT IT TESTS: quantifying upside before investing. OUTLINE: locate the affected funnel step, estimate addressable population times a bounded conversion lift times value per user, then sanity-check against a realistic ceiling.

Growth & Experimentation2 min read

Combine qualitative and quantitative data for hypotheses

WHAT IT TESTS: mixed-methods reasoning to build strong hypotheses. OUTLINE: quant reveals what and where, qual reveals why, then triangulate into a falsifiable hypothesis with a metric. RED FLAG: treating anecdotes as proof or analytics as self-explanatory.

Growth & Experimentation2 min read

Explain RICE scoring and its Confidence factor

WHAT IT TESTS: understanding RICE and the role of Confidence. OUTLINE: score equals Reach times Impact times Confidence divided by Effort; Confidence discounts uncertain estimates; ground it in evidence tiers.

Growth & Experimentation2 min read

Instrument a first-full-song activation event

WHAT IT TESTS: precise event definition and reliable instrumentation. OUTLINE: define 'full song' server-side, emit a typed event with user, song, and context, dedupe the first-time flag.

Growth & Experimentation89 sec read

Architect an experimentation dashboard for culture

WHAT IT TESTS: product thinking about experimentation as an organizational system, not just stats. OUTLINE: searchable experiment repository, structured hypotheses, results regardless of outcome, and cross-team discovery.

Growth & Experimentation85 sec read

Resurrection Campaign

A resurrection campaign is a targeted effort to win back dormant or churned users by re-engaging them with relevant value, often via email or push. It matters because reactivating known users is usually cheaper than acquiring new ones.