Cross-border commerce, personal care appliancesProduct category：Electric nose and ear hair trimmer
The customers you think are good are mostly heading for the door
Value and lifecycle diagnosis across 11,627 customers. Acquisition looks strong; underneath, the foundation is loosening.
- Engine used
- TagPilot
- Report date
- 2026-07
The question
Of these 11,627 customers, how many will still be here tomorrow?
The situation
In this market the brand faces pressure from two directions at once: a base that has not yet set, and a core that is coming apart. Revenue itself shows nothing unusual, which is exactly why an ordinary report cannot surface the problem.
Key numbers
- 11,627位
- Customers analysed
- 46.7%
- Share of CLV from high-value customers
- 73.8%
- High-value customers in the risk zone
What the analysis found
- 01
Value is dangerously concentrated
High-value customers are 22.1% of the base but contribute 46.7% of total CLV. Revenue leans on a small group; when they move, the whole revenue curve moves with them.
- 02
Three quarters of the high-value base is already loosening
73.8% of high-value customers sit in the high CLV, low survival-probability zone. They have not formally left, but repurchase probability is collapsing. Financials will not show it for a while, and by the time it is felt it is too late to save them.
- 03
Repeat purchase has not formed as a habit
Average purchase frequency is 1.52, and mean CLV runs 54% above the median. A few very high-frequency customers pull the average up while most of the base stops at one or two orders.
Marketing strategy you can execute
- Hold the high-value base and prevent churn. This carries the highest return: winning back one high-value customer costs far less than acquiring an equivalent new one.
- Grow new customers into core customers. New customers are 74.5% of the base, the largest pool, yet core-customer order value sits 0.8% below new-customer order value. The upgrade path has effectively stopped working, so growth lies in buying more often, not in buying more expensively.
- Reactivate dormant and lapsed customers. Small in scale but not to be written off. Diagnose before spending: without knowing why they left, a discount buys one order and nothing more.
About these numbers
The above are diagnoses and estimates produced by our models from existing data. They support marketing decisions, they are not performance guarantees, and they are not results measured after adoption. Client brands, competitor names and product identifiers have been removed.
Start by seeing what your own data looks like
A thirty-minute walkthrough on your data, not a canned demo.
Book a demo