Onboarding Activation

Onboarding & Activation

Get new users to the moment the product genuinely works for them — the point after which they're substantially more likely to stay. Everything before that moment is cost; everything after is compounding.

Activation is where most retention problems actually live. When a cohort curve drops steeply in the first weeks and then flattens, that is almost never a price problem or a competitor problem. Those users never reached value, so there was nothing to retain. Discounting them back does nothing, because the product never worked for them in the first place.

Two things make this hard. The activation moment is usually assumed rather than measured — teams pick something plausible like "completed onboarding" that doesn't actually predict retention. And onboarding is designed as a tour of the product rather than a path to one outcome, so users learn what exists without ever getting anything.

When to use

The user is improving product onboarding, first-run experience, or activation rate, or asking why signups don't convert to active users. For diagnosing whether activation is the problem, run lifecycle-and-retention first — its cohort view is what identifies a steep early drop. For the flow before the account exists, use signup-flow-optimizer.

Before you start

1. Read the brand/product context (.agents/product-marketing.md) for the audience and what the product is for.
2. Get cohort retention if it exists. A steep early drop confirms activation is the problem; a gradual decline points elsewhere and this skill is the wrong tool.
3. Ask what retained users did that churned users didn't — in their first week. That's the raw material for defining the activation moment properly.
4. Walk the actual first run. From account creation to the first useful outcome, on the real product, as a new user with no data.

Method

Define the activation moment from evidence, not intuition. The right definition is a specific, countable action — or small set of actions — where users who complete it retain materially better than users who don't. Compare retention for users who did versus didn't do each candidate action in week one. The one with the widest gap is your activation moment.

Good definitions are concrete: imported first customer list, invited a teammate, created the second project, ran the first report. Bad definitions are proxies for effort rather than value: completed the tour, logged in three times, finished onboarding.

A correlation caveat worth stating. Users who invite a teammate retain better, but inviting isn't necessarily what causes retention — engaged users do both. Treat the activation moment as the best available leading indicator, and validate by testing whether driving the action moves retention. If it doesn't, you found a symptom.

Then find where users stall. Instrument each step from signup to activation and measure drop-off per step. The single steepest drop is where the work goes — and it's frequently earlier and more mundane than teams expect: an empty state with nothing to do, a data import that requires a CSV nobody has to hand, a permission the user can't grant themselves.

Then shorten the path. Every step between signup and value is a chance to lose someone. The goal is not a better tour; it's fewer steps to the same outcome.

See resources/activation-playbook.md for defining the moment, the stall patterns, first-run design patterns, and a worked example.

The math lives elsewhere

This skill deliberately ships no engine. The retention and LTV consequences of a change in activation rate are already computed by lifecycle-and-retention, which models churn, cohort curves, and the lifetime effect. Duplicating that here would give you two engines that could disagree.

Use them together: lifecycle-and-retention tells you activation is the problem and what fixing it is worth; this skill fixes it.

Workflow

1. Confirm activation is the problem. Steep early cohort drop, not a gradual decline. If it's gradual, this is the wrong skill.
2. Define the activation moment from data — the week-one action with the widest retention gap between those who did and didn't.
3. Sanity-check it. Is it plausible that this action causes stickiness, or is it just what engaged users happen to do? Note the uncertainty either way.
4. Map the path from account creation to that moment, step by step, as it exists.
5. Instrument every step and find the steepest drop.
6. Attack the steepest stall first using the patterns in the playbook — usually by removing a step, pre-filling data, or replacing an empty state.
7. Shorten the path overall. Count the steps before and after; if the number hasn't fallen, the redesign is decoration.
8. Specify the new first run concretely enough to build, including empty states and the failure paths.
9. Instrument activation as a metric — rate and median time-to-activation, tracked by cohort so improvements are visible.

Present the result

Lead with the activation moment and the evidence for it — the retention gap that justifies the definition. That single decision governs everything downstream, and getting it wrong makes every subsequent optimisation point the wrong way.

Then the stall map with the steepest drop named, then the redesigned first run, then the metric to instrument. Flag explicitly where you've inferred the activation moment rather than measured it.

Guardrails & common mistakes

• Don't assume the activation moment. The most common failure is optimising toward something that feels important and doesn't predict retention. Measure the gap.
• Correlation isn't causation here. Validate by driving the action and watching whether retention moves.
• Onboarding is not a product tour. Tours teach the interface; activation delivers an outcome. A user who has seen every feature and accomplished nothing is not activated.
• Empty states are where activation dies. A new account with no data and no obvious first move is the most common stall. Sample data, templates, or a guided first object usually beat any amount of instructional copy.
• Time-to-value matters as much as activation rate. The same rate reached in one day rather than one week retains better and costs less to support.
• Don't make onboarding mandatory. Skippable, resumable, and never blocking. Forced flows produce completion, not activation.
• In-product beats email. Email supports activation; it can't carry it. If your onboarding emails are doing the work, the product's first run is the actual problem — see email-lifecycle-sequence for the supporting half.
• Segment where it matters. Different roles or use cases often have genuinely different activation moments. One path for all of them under-serves each.
• If activation is fine and retention still falls later, the problem is value delivery over time, not onboarding. Go back to lifecycle-and-retention.

Related AAJ resources

• Interactive tool: https://aajconsult.com/tools/churn-nrr-calculator
• Template: https://aajconsult.com/resources/retention-nrr-workbook

Related skills

lifecycle-and-retention (diagnose first; it owns the retention and LTV math) · signup-flow-optimizer (the flow before the account exists) · email-lifecycle-sequence (the email that supports activation) · marketing-psychology (the friction blocking the first action) · customer-survey-design (ask stalled users what stopped them).

Credits

Original AAJ skill. The Agent Skills format and the marketing-skills catalog by Corey Haines (coreyhaines31/marketingskills, MIT) were references for structure and coverage; this skill is independently written. See the repository README for the full reference list.