Introduction

Most DTC merchants know exactly what they spent to acquire a customer last month. Far fewer know how many were still buying 60 days later. When repeat-purchase cohorts weaken, CAC climbs and forecasts miss long before revenue shows it.

Aggregated H1 2026 fulfillment data across cross-border DTC stores shows merchants running quarterly cohort audits retain 1.8× more second-order revenue than stores reviewing only top-line sales. This article covers four silent churn signals, a five-step audit workflow, and re-engagement thresholds — critical before Q4 traffic costs peak.

Industry Benchmark: Repeat-Purchase Cohort Performance Tiers

We analyzed order records from cross-border DTC stores in H1 2026, sorting merchants into three tiers by 60-day second-order and 120-day third-order cohort behavior:

  • High-retention stores (Top 22%): 28%–35% of first-time buyers reorder within 60 days; 12%+ reach a third order within 120 days.
  • Mid-tier stores (44%): 60-day second-order rate of 15%–24%; retention leans on promotions, eroding margin with each reactivation (see our Discount Erosion Audit).
  • Low-retention stores (Bottom 34%): Under 12% return within 60 days. These stores are customer-replacement machines — every growth dollar must be re-bought with ad spend.

The pattern separating the tiers: high-retention stores detect cohort decay early and intervene while re-engagement is still cheap. Retention is won or lost in the 30–60 day window after first purchase — long before a customer looks "lost."

Four Silent Churn Signals That Appear Before Customers Disappear

  • Signal 1: Email engagement decay. Open rates falling below 15% across three campaigns — after 30%+ engagement — cut 60-day repeat probability to one-third. The subscriber is still on the list but already gone in practice.
  • Signal 2: Time-to-second-order drift. When median time-to-second-order runs 50%+ past the category baseline (e.g., 38 → 61 days vs. a 35-day norm), the cohort is decaying even if repeat counts look stable.
  • Signal 3: Unresolved support tickets. Tickets resolved within 24 hours correlate with nearly 2× repeat rates; unresolved past 48 hours, 40% below baseline. Silence after a bad experience is departure without announcement.
  • Signal 4: First-order discount dependence. Cohorts acquired with 30%+ first-order discounts repeat 35%–50% less than full-price cohorts — and only respond to deeper discounts.

Store Case: Reading the Cliff Early

A home & kitchen DTC store with $180K–$180K–$220K monthly revenue held stable sales through Q1–Q2 2026 while net margin compressed from 14% to 8%. A cohort audit found the cause: 60-day second-order rate had slid from 24% to 14% over two quarters while CAC climbed 22% — invisible in monthly revenue charts.

Three failures were running at once: email engagement decay triggered 41 days after first purchase, median time-to-second-order stretched from 38 to 61 days against a 35-day baseline, and 58% of new customers came from 35%-off influencer codes.

The store rebuilt retention around the 30–60 day window: engagement-based win-back triggers, a post-delivery care sequence, and influencer codes repriced to protect full-margin second orders. A quarter later, second-order rate recovered to 21% and blended CAC fell back — with no new traffic budget. Reliable 2–4 day fulfillment kept post-purchase sentiment positive throughout.

Step-by-Step Repeat-Purchase Cohort Audit Workflow

Run quarterly (monthly if under $100K revenue). Requires only an order export, spreadsheet pivots, and email engagement data.

  1. Build monthly first-order cohorts for the trailing 12 months. Calculate second-order rates at 30/60 days and third-order within 120 days.
  2. Overlay engagement velocity. Flag cohorts where median email engagement drops below 15% before day 60.
  3. Measure time-to-second-order drift. A 15+ day drift past category baseline is a warning; 50%+ means active decay.
  4. Attribute acquisition source per cohort. Split by channel and discount depth to see which channels deliver returning customers — and which deliver one-time bargain hunters.
  5. Intervene at the signal, not the symptom. Win-back flows at engagement decay (day 30–45), care sequences at delivery, channel repricing at acquisition. Re-audit after 30 days to confirm recovery.

Core Takeaways

  • Retention is decided in the 30–60 day post-purchase window, long before customers show up as "lost" in monthly revenue.
  • The four highest-fidelity churn signals: email engagement decay, time-to-second-order drift, unresolved support tickets, and first-order discount dependence.
  • Revenue stability can hide cohort collapse; only cohort-level audits reveal it in time to act cheaply.
  • Discount-acquired customers repeat 35%–50% less — judge channels on 60-day cohort value.
  • Quarterly cohort audits plus fast, reliable fulfillment are the cheapest retention system a DTC store can run in H2 2026.