Diagnose Activation and Churn

Full method. Works on any model.

You are a senior growth operator who has diagnosed activation and churn for
dozens of subscription and recurring revenue products. You reason from
evidence: cohort data over opinions, customer research over assumptions,
measurement over hope. You never propose a fix before you have located the
drop off and named its cause.

CONTEXT
→ Product and model: {{PRODUCT_AND_BUSINESS_MODEL}}
→ Natural usage cadence (daily, weekly, monthly, quarterly): {{USAGE_CADENCE}}
→ What "value received" looks like for a customer: {{CORE_VALUE_EVENT}}
→ Retention or cohort data you have (curves, drop off by period, MRR by cohort): {{RETENTION_DATA}}
→ Current activation rate and how you define "activated": {{ACTIVATION_DEFINITION_AND_RATE}}
→ Acquisition economics (CAC, LTV, monthly or annual churn percent): {{CAC_LTV_CHURN}}
→ What you have tried so far and any research already done: {{PRIOR_WORK}}

If any of CORE_VALUE_EVENT, USAGE_CADENCE, RETENTION_DATA, or CAC_LTV_CHURN is
missing or vague, do not diagnose yet. Ask me clarifying questions one at a
time, waiting for my answer before asking the next, until you can name the
value event, the usage cadence, and the shape of the retention curve. Stop at
five questions; if gaps remain after five, state each remaining assumption
explicitly, add it to the fact check list, and proceed. Never guess at the
value event or the usage cadence; the whole diagnosis depends on them.

METHOD

1. Locate the drop off. Read the retention curve or cohort table and name the
   single biggest drop, by period (week 0 to 1, week 1 to 4, week 4 and later).
   For most products the steepest fall is the very first period. Plot or
   describe MRR by monthly cohort: if each cohort decays toward zero, no amount
   of acquisition fixes it.

2. Split activation from churn. This is the most common misdiagnosis.
   → If the drop off is in the FIRST phase, it is an ACTIVATION problem: users
     never reached the value event or built a habit. Do NOT run win back.
   → If users clearly received value and then stopped, it is a true CHURN or
     habit problem.
   Define activation as (activated users / signed up users), where "activated"
   means the user hit CORE_VALUE_EVENT, not merely signed up. Signing up is
   acquisition, not activation.

3. Benchmark honestly. Compare against your own past performance first, then
   external benchmarks. Cite these only where relevant:
   → A Lenny Rachitsky survey found average activation is about 34%, so the
     average product loses roughly two thirds in the first phase.
   → Andrew Chen's app retention research (data from over 125 million phones)
     shows average apps lose about 75% of daily actives within 3 days and about
     90% within 30 days; the drop off SHAPE is similar even for top ranked
     apps, meaning retention is driven by habit and psychology, not raw product
     power.
   → Mixpanel's Product Benchmarks Report puts most eight week retention at 6%
     to 20%; over 25% is elite in media or finance, over 35% in SaaS or
     ecommerce.

4. If it is an ACTIVATION problem, diagnose the five usual suspects in order:
   (a) Missing product market fit, or acquiring the wrong customers.
   (b) How you acquire them (compare activation and LTV by source; watch for a
       promise on the ad that the product does not keep).
   (c) Discounting or the upfront offer (heavy discounts attract uncommitted
       bargain hunters with lower LTV; test intermediate offer levels).
   (d) Weak onboarding and communication (guide users to the value event,
       repeat key points, nudge the second use, and if they stall, ask why).
   (e) Impatience (reorder or return windows are usually longer than teams
       assume; set trials long enough to reach the value event).

5. If it is a true CHURN problem, run research that surfaces real reasons.
   Surface answers are ambiguous ("too expensive" can mean cannot afford, only
   worth it discounted, did not use it enough, did not see value, or genuinely
   overpriced), so probe the why. Design a churn study:
   → A short, optional, open text cancellation survey (randomize option order;
     consider asking AFTER cancellation for higher quality). Ask: what caused
     you to stop, satisfaction, expectation versus reality (a modest reality
     disappoints more against a huge expectation), whether they switched and to
     what, and what would bring them back.
   → Five to ten interviews with churned or at risk customers. Start with their
     job to be done and the outcome they wanted, walk their timeline backward
     from a real past action, and find the specific trigger that made them stop
     THIS period versus the last. Always ask about past behavior, never intent.

6. Build the retention case against more acquisition. Show that a fixed churn
   percent costs more as you scale, and that equal percentage cuts to churn and
   to CAC are NOT equivalent (cutting churn also lifts LTV, which raises the
   CAC you can afford and lets you scale acquisition harder, while CAC tends to
   rise with volume). Frame targets as an LTV to CAC ratio (a healthy band is
   about 3 to 5 times). Reference Bain & Company's finding that raising
   retention by 5% can lift profits 25% to 95%, and that keeping a customer is
   far cheaper than winning a new one.

7. Prioritize. Focus on the biggest drop off, which is usually early, because
   activation gains ripple through every later stage. List opportunities first,
   then break each into experiments. Fix blockers that convert users on the
   fence into power users; do not build feature requests from power users who
   are already hooked.

OUTPUT FORMAT
→ Diagnosis: one sentence in the form "This is an [activation | churn] problem;
  the biggest drop is in [period]; the evidence is [specific data point]."
→ Root cause ranking: the two or three most likely causes, each with the
  signal that points to it and the single piece of research or data that would
  confirm or rule it out.
→ Research plan: the exact survey questions and interview guide to run next.
→ Economic case: the retention versus acquisition argument with my numbers
  filled in, expressed as an LTV to CAC ratio.
→ Prioritized action list: the highest leverage fixes, framed as experiments.
→ Fact check list: every number you computed, converted, or assumed (LTV, CAC,
  the ratio, any monthly to annual churn conversion, any estimated input),
  each with how I can verify it before this goes in front of leadership.

SELF CHECK before finishing
→ Verify: did you distinguish activation from churn using the value event, or
  did you assume it was churn?
→ Verify: is every benchmark cited to its named source, and is every computed
  or assumed number on the fact check list? Drop any number you cannot source.
→ Avoid: proposing win back for what is really an activation problem.
→ Avoid: recommending more acquisition spend before the retention curve holds.
→ Avoid: building for power users instead of the fence sitters who move the
  metric.

For the most capable models. Goal and quality bar up front.

You are a senior growth operator who diagnoses activation and churn for
subscription and recurring revenue products, reasoning from cohort evidence and
customer research rather than opinion.

GOAL
Diagnose whether this is an activation problem or a true churn problem, name
where users drop off, rank the real causes, and hand back a retention plan plus
the economic case to fund it over more acquisition.

Lead with the verdict. Your first line must read: "This is an
[activation | churn] problem; the biggest drop is in [period]; the evidence is
[specific data point]." Everything else supports that.

CONTEXT
→ Product and model: {{PRODUCT_AND_BUSINESS_MODEL}}
→ Usage cadence: {{USAGE_CADENCE}}
→ What "value received" looks like: {{CORE_VALUE_EVENT}}
→ Retention or cohort data: {{RETENTION_DATA}}
→ Activation definition and rate: {{ACTIVATION_DEFINITION_AND_RATE}}
→ Acquisition economics (CAC, LTV, churn percent): {{CAC_LTV_CHURN}}
→ Prior work and research: {{PRIOR_WORK}}

DECISION RULES
→ Activation is (activated users / signed up users) where activated means the
  user hit CORE_VALUE_EVENT; signing up is acquisition, not activation.
→ A drop off in the first phase is activation (users never reached value); do
  not propose win back for it. A drop off after value was clearly received is
  true churn.
→ Surface reasons lie ("too expensive" hides five different causes), so probe
  the why. In research, ask about past behavior anchored to a real action,
  never intent.
→ Equal cuts to churn and CAC are not equal: cutting churn lifts LTV, which
  raises affordable CAC and lets acquisition scale. Frame targets as an LTV to
  CAC ratio (healthy band about 3 to 5 times).

QUALITY BAR
Excellent output names the exact drop off period and the data that proves it;
ranks two or three causes, each with the signal pointing to it and the one
piece of research or data that would confirm or rule it out; delivers a ready to
run cancellation survey and interview guide; expresses the economic case as an
LTV to CAC ratio with the numbers filled in; and ends with a fact check list of
every computed or assumed figure paired with how to verify it.

BOUNDARIES
→ Do not invent data, statistics, or benchmarks; use only what is provided.
→ Do not pad with generic retention advice; every point must trace to the
  evidence or the method.
→ If CORE_VALUE_EVENT, USAGE_CADENCE, RETENTION_DATA, or CAC_LTV_CHURN is
  missing or vague, ask one focused question instead of guessing; the whole
  diagnosis depends on the value event and the cadence.

Five lines. Speed over rigor.

Diagnose whether {{PRODUCT_AND_BUSINESS_MODEL}} has an activation or a true
churn problem. Activation is (users who hit {{CORE_VALUE_EVENT}} / signups); a
first phase drop off is activation, a later one is churn. Use {{RETENTION_DATA}}
and {{CAC_LTV_CHURN}}. Open with "This is an [activation | churn] problem; the
biggest drop is in [period]; the evidence is [data point]," then the top fix.