When to use
- You want to know what a customer is worth over time, not just at first purchase — and whether that's getting better or worse.
- You can share orders with a customer id, order date, and amount (a transactions export is perfect).
- You want to see retention and cumulative value by cohort, and understand what it means for how much you can spend to get a customer.
Do not use it to invent lifetime value, retention curves, or future revenue. If the history is too short to project, the skill says so and sticks to what actually happened.
What it does
Runs a cohort analysis of customer value:
- Builds cohorts by first-purchase month (or another period you choose) from your transactions.
- Tracks retention: what share of each cohort is still buying 1, 3, 6, 12 months in.
- Tracks cumulative value: how much an average customer from each cohort has spent by each month — the real LTV curve, built from your data.
- Compares cohorts: are newer customers spending more or less, staying longer or shorter, than older ones — and calls out the trend.
- Explains what it means: a rough sense of how much you can afford to spend to acquire a customer, with the caveat that it depends on margin and payback period.
- Flags the limits: short history, thin cohorts, or missing months — it labels projections as assumptions, never as facts.
- Closes with a verdict (always): a 0–100 score for how trustworthy this analysis is given your data depth; a traffic light (🟢 80+ decision-ready; 🟡 50–79 directional, get more history; 🔴 <50 too little data to trust the curves); the single insight that matters most; and a note that the score reflects your data, not a guarantee of future value.
How to use
- Share your transactions: customer id, order date, and amount per order.
- Ask: "Do an LTV cohort analysis by first-purchase month and tell me if my newer customers are worth more."
- Review the cohorts and the trend; ask it to show its inputs if a number surprises you.
- Use the payback view to sanity-check what you spend on acquisition — with your own margin.
Best for
Ecommerce, subscriptions, and repeat-purchase businesses that want to understand real customer value and how much they can spend to grow. Works best with at least several months of transaction history.