Customer segmentation and reactivation opportunities · PetMarket
Example business: PetMarket — a pet food and accessories e-commerce with recurring purchases.
The ten sections organize this illustrative example. Each product report has its own scope and format.
Analysis overview
Period, cut-off date and customer population.
- Business
- PetMarket (fictitious data)
- Analyzed period
- July 2023 – June 2025 (24 months)
- Cut-off date
- June 30, 2025
- Population
- 8,420 customers with at least one purchase in 24 months.
Executive summary
Composition of the base and main opportunities.
The base was split into five segments by recency, frequency and value. Two segments — VIP and Recurring — are 33% of customers and 70% of value. The main risk lies in the 1,730 inactive customers and in the 63% of new customers who do not repurchase within their first 90 days.
- 610 VIP customers generate 31% of value with a USD 68 average ticket.
- Inactive customers were worth USD 48,700 in their last active year: that is the reactivation pool.
- The actionable list in the annex assigns a segment to every customer by identifier.
Quality and customer identification
Identifier coverage and history limits.
96% of purchases carry a customer identifier; the remaining 4% are guest checkouts and are excluded from the segmentation. The 24-month history is enough to observe recurrence in a category with a 4-to-6-week purchase cycle.
- 26,090 orders with an identified customer.
- 214 duplicate customers merged by equivalent email.
- The annex list uses internal identifiers only, with no personal data.
Key indicators
Customers, frequency, recency and purchase value.
Active customers (12 months)
6,690
+7%
Average annual frequency
3.1
Average recency
74
Days since last purchase
Average value per customer per year
USD 96
Segmentation
Criteria used, size and contribution of each group.
Segments were built with the recency, frequency and value rules defined in the methodology; every customer belongs to exactly one segment.
Value distribution by segment
Each segment’s share of total 24-month value.
- VIPUSD 214,80031%
- RecurringUSD 268,30039%
- NewUSD 84,60012%
- OccasionalUSD 76,10011%
- InactiveUSD 48,7007%
Explore with the pointer or keyboard. Tap a value to keep its detail visible.
Purchase behavior
Repurchase, cohorts and inactivity signals.
The 90-day repurchase rate of new customers has been falling cohort by cohort: from 41% in the last quarter of 2024 to 34% in the first quarter of 2025. The inactivity signal is defined as 120 days without a purchase, twice the typical cycle.
90-day repurchase by first-order cohort
Share of each quarter’s new customers who purchased again within 90 days.
Explore with the pointer or keyboard. Tap a value to keep its detail visible.
Actionable list
Customer identifiers and assigned segment, without unnecessary personal data.
The annex includes the full 8,420-row actionable list. This excerpt shows the format: internal identifier, assigned segment and the variables that justify it. It contains no names, emails or phone numbers.
AI analysis
Differentiated strategies per segment and hypotheses for commercial tests.
Key findings
Value is concentrated and stable
VIP and Recurring concentrate 70% of value and their average recency (21 and 38 days) shows they remain active. The current revenue base is not at immediate risk.
Evidence: Segment table (element 5).
The leak is in the first repurchase
63% of new customers do not repurchase within 90 days, and recent cohorts repurchase less than those of a year ago. Acquiring customers is working; converting them into recurring ones is not.
Evidence: Repurchase-by-cohort chart (element 6).
The reactivation pool has a known value
The 1,730 inactive customers generated USD 48,700 in their last active year. 640 of them were Recurring or VIP before going quiet: they are the subgroup with the most reasonable response likelihood.
Evidence: Annex actionable list, filtered by previous segment.
Business interpretation
Each segment calls for a different strategy: protect VIPs with retention benefits, automate the new customers’ second purchase and reactivate former recurring customers first. Sending the same message to the whole base would dilute budget on the 2,010 occasional customers, the segment with the lowest expected response.
Hypotheses and limitations
- The 2025 cohorts’ repurchase decline coincides with January’s shipping-fee increase; a free-shipping-on-second-purchase test is needed to validate it.
- Some inactive customers may have changed pets or cities; that data does not exist in the file and caps reactivation expectations.
Prioritized recommendations
Automate the second-purchase journey
A two-message series between days 25 and 40 after the first order, with a food-cycle reminder. This lever attacks the largest measured leak.
Staged reactivation campaign
Start with the 640 former recurring customers with a comeback offer; extend to the remaining inactive customers only if the first wave clears the campaign’s break-even.
VIP benefits program
Priority shipping or a subscription discount for the 610 VIPs. The goal is to defend the 31% of value, not to grow it.
Proposed follow-up
Campaign plan
Actions, priorities and repurchase or reactivation metrics.
Methodology and annexes
Segmentation rules, reference date and detailed results.
Rule-based segmentation on recency, frequency and value as of June 30, 2025: VIP (high value and active), Recurring (3+ purchases and active), New (first purchase within 120 days), Inactive (more than 120 days without buying), Occasional (the rest). Exact rules and thresholds are in the annex.
- Annex A: full actionable list (8,420 rows, identifiers only).
- Annex B: repurchase matrix by monthly cohort.
- Annex C: segmentation rule thresholds.