Retention Funnels
Retention funnels show how customer cohorts decay over time — Day 1, Day 7, Day 30, Day 90. The shape of that curve predicts long-term LTV more reliably than first-order conversion rate.
Retention Funnels
A view of how a customer cohort decays over time — Day 1, Day 7, Day 30, Day 90 — used to predict LTV.
A retention funnel tracks what percentage of a customer cohort is still active, purchasing, or engaged at fixed intervals after their first session or first order. Unlike a conversion funnel, which measures one-time movement through checkout, the retention funnel measures repeated behaviour: did the Day-1 buyer come back on Day 7? Did they reorder by Day 30? Did they still exist on Day 90?
The shape of this decay curve is the single most reliable predictor of long-term LTV. Two stores with identical first-order conversion rates can have wildly different valuations if one flattens at 22% by Day 90 and the other bleeds to 6%.
Retention funnels sit inside the broader practice of funnel analytics, but they answer a different question. Conversion funnels ask "how many visitors became buyers this week?" Retention funnels ask "of the buyers we acquired in March, how many are still buying in June?"
The reason this matters: paid acquisition costs are rising, and a store that retains 25% of buyers at Day 90 can spend roughly three times more on acquisition than one retaining 8% — for the same payback window. The retention curve is what turns CAC math from a guess into a forecast.
Retention_N = Active_at_day_N / Cohort_size_at_day_0
Retention_N
N-day retention rate
Percentage of the original cohort still active or purchasing at day N
Active_at_day_N
Active cohort members at day N
Number of customers from the original cohort who placed an order, opened the app, or returned to the site by day N
Cohort_size_at_day_0
Original cohort size
Total customers acquired in the cohort window (typically a single week or month of first orders)
A skincare brand acquires 1,200 first-time buyers in March. Tracking the cohort, 168 of them place a second order within 90 days.
Active at Day 90: 168 customers
Cohort size at Day 0: 1,200 customers
→ Retention_90 = 14%
14% Day-90 retention is healthy for a skincare SKU on a 60-day replenishment cycle. If the same brand sold a one-shot category like luggage, 14% would be exceptional; if it sold a daily-use consumable, 14% would suggest the post-purchase flow is leaking buyers to competitors.
Read retention curves by their shape, not their endpoint. A curve that crashes between Day 1 and Day 7 then plateaus is a different problem from one that decays steadily across 90 days. The first usually points to product-experience or expectation mismatch; the second points to lifecycle marketing and replenishment timing.
Typical Day-30 and Day-90 retention rates by DTC vertical (repeat-purchase rate of first-time buyers).
| Vertical | Day-30 retention | Day-90 retention | Replenishment cycle |
|---|---|---|---|
| Beauty & skincare | 12-18% | 22-32% | 30-60 days |
| Apparel & accessories | 6-10% | 14-20% | 60-120 days |
| Supplements & wellness | 18-28% | 35-45% | 30 days (subscription-led) |
| Home & kitchen | 3-6% | 8-14% | 90-180 days |
| Pet food & treats | 22-32% | 40-55% | 30-45 days |
| Consumer electronics | 2-4% | 5-9% | 180+ days |
A flat tail matters more than a high peak. If your Day-7 to Day-90 line is roughly horizontal, you have a real repeat-buyer base and a forecastable LTV. If it keeps sloping down past Day 90, you are renting customers from paid channels — every cohort needs replacing in full.
Retention funnel FAQ
A conversion funnel measures one-time movement through a single journey (visit → product page → cart → checkout). A retention funnel measures repeated behaviour across time (Day 1 → Day 7 → Day 30 → Day 90) for a fixed cohort. Both live under funnel analytics, but they answer different questions: "did we win the sale?" versus "did we keep the customer?"
It depends on your replenishment cycle. For supplements and pet food on 30-day cycles, Day-30 is the leading indicator. For apparel on 90+ day cycles, Day-90 is where the signal lives. For app-based or content-led brands, Day-1 and Day-7 are the early-warning signals. Pick the day that lines up with your natural repurchase window.
Start with order data exported from Shopify or WooCommerce. Group customers by their first-order month, then count how many of each cohort placed another order within 7, 30, and 90 days. A spreadsheet works for the first version. Tools that import historical GA4 and order data can produce the cohort grid automatically, so you don't wait six months to see your first curve.
AOV tells you what a buyer spends once. The retention curve tells you how many times they'll spend. Two stores with identical €70 AOV but Day-90 retention of 8% versus 25% have completely different 12-month LTVs — roughly three times apart — even before you factor in margin compounding from organic repeat orders.
It's vertical-specific. Skincare and supplements healthy zone is 25-40%; apparel is 14-20%; durables like home and electronics live below 15% by design. Compare against your category, not against "DTC" as a whole. Comparing a luggage brand to a vitamin subscription is meaningless.
If you're starting from zero data, 90-120 days for your first full curve. If you import historical GA4 and order history, you can produce a Day-90 cohort grid on day one — which is the difference between waiting a quarter to diagnose retention and starting to fix it this week.
Both, and they tell different stories. Customer retention (how many buyers came back) is the leading indicator of brand health. Revenue retention (how much they spent on the second and third order) reveals expansion versus contraction within the retained base. A flat customer curve with rising revenue per retained buyer is a healthier picture than the reverse.
Churn is typically a subscription concept — a binary cancellation event. Retention funnels work for transactional commerce where there is no formal cancellation, only an absence of repeat orders. You define a "churned" buyer as one who hasn't reordered within N days, where N is roughly 2-3x your replenishment cycle.
Yes, and you should. Buyers acquired via paid social often retain at half the rate of buyers acquired via organic search or referral. Layering channel onto the cohort grid is how you spot when your blended CAC math is masking a structurally unprofitable channel — the Day-90 curve will tell you before the P&L does.
Look at the gap between Day-1 and Day-30. If buyers are leaving in the first week, the issue is usually product expectation, packaging, or onboarding email timing. If they show up at Day-30 but vanish by Day-60, the issue is replenishment trigger — they ran out, no one reminded them, and they bought a competitor's product on autopilot.
Get an AI expert review of your site
Paste your URL — Metricuno's AI runs the same heuristic checks a senior CRO consultant would, scoring your page and prioritising the fixes that'll move conversion fastest.