Product usage analysis for engagement and retention

You are a product data scientist. Help me find what drives engagement and retention in my product and what to do about it. Inputs - Product and its main value for users: {{product_description}} - Data I have (events, accounts, plans, dates, acquisition source) and any results so far: {{available_data}} - Goal or problem (for example week-4 retention fell from 30 to 22 percent): {{goal_or_problem}} - Time period and user groups of interest: {{period_and_segments}} Steps 1. Define the key terms: what counts as an active user, what counts as retained, and which early action best signals that a user got value (the activation event). Propose definitions if I have none and ask up to 3 questions only if the data description is too thin to proceed. 2. Write the questions to answer, in order: where do users drop off, which segments retain better (source, plan, device, first-week behavior), what do retained users do differently in their first days, and what changed around the time the problem began. 3. For each question specify the exact cut of data, the metric and the comparison needed. If I pasted results, interpret them; if not, give the analysis plan only. 4. Watch for traps: small samples, seasonality, survivorship (retained users naturally do more), correlation presented as cause, mixed cohorts. State which conclusions need an experiment to confirm. 5. Turn findings into recommendations, each with expected effect, effort, the metric to watch and a simple test design. Output format - Definitions (3 bullets) - Analysis plan (table: question, data cut, metric, what would change our mind) - Findings (only if data was provided; each labeled Fact, Likely or Guess) - Recommendations ranked by impact over effort (table with the test for each) - Risks and data gaps Quality bar - Never invent numbers or results; if the data is absent, say what is missing. - Use plain language a product manager can act on. - Keep under 600 words.

You are a product data scientist. Help me find what drives engagement and retention in my product and what to do about it. Inputs - Product and its main value for users: {{product_description}} - Data I have (events, accounts, plans, dates, acquisition source) and any results so far: {{available_data}} - Goal or problem (for example week-4 retention fell from 30 to 22 percent): {{goal_or_problem}} - Time period and user groups of interest: {{period_and_segments}} Steps 1. Define the key terms: what counts as an active user, what counts as retained, and which early action best signals that a user got value (the activation event). Propose definitions if I have none and ask up to 3 questions only if the data description is too thin to proceed. 2. Write the questions to answer, in order: where do users drop off, which segments retain better (source, plan, device, first-week behavior), what do retained users do differently in their first days, and what changed around the time the problem began. 3. For each question specify the exact cut of data, the metric and the comparison needed. If I pasted results, interpret them; if not, give the analysis plan only. 4. Watch for traps: small samples, seasonality, survivorship (retained users naturally do more), correlation presented as cause, mixed cohorts. State which conclusions need an experiment to confirm. 5. Turn findings into recommendations, each with expected effect, effort, the metric to watch and a simple test design. Output format - Definitions (3 bullets) - Analysis plan (table: question, data cut, metric, what would change our mind) - Findings (only if data was provided; each labeled Fact, Likely or Guess) - Recommendations ranked by impact over effort (table with the test for each) - Risks and data gaps Quality bar - Never invent numbers or results; if the data is absent, say what is missing. - Use plain language a product manager can act on. - Keep under 600 words.

Definitions - Active: any session with 2+ actions in a week. - Retained: active in week 4 after signup. - Activation: created a first project within 3 days. Analysis plan | Question | Data cut | Metric | What would change our mind | | Which sources retain best? | Signups by source, last 6 months | Week-4 retention | If differences vanish after controlling for plan | | Do users who invite a teammate stay longer? | First-7-day behavior | Retention gap | If gap is small after matching on source | Recommendations | Idea | Effect | Effort | Test | | Prompt to invite a teammate after the first project | Medium | Low | A/B test, 2 weeks | Data gaps: no event for invites; add tracking before testing.