Retention Thresholds Worth Writing Down

Repeat Purchase Rate (RPR) - The Core Metric

Repeat purchase rate is the share of customers from a cohort who buy again within a set window. For DTC Shopify brands, the standard window is 90 days post-first purchase. A customer acquired on January 1 who buys again by March 31 counts as retained.

Calculation: (Customers with 2+ purchases in window / Total customers in cohort) × 100.

Benchmark thresholds by category:

- Consumables (food, supplements, beauty): 25-40% RPR at 90 days is acceptable; 40%+ is strong.

- Apparel / discretionary: 12-20% RPR at 90 days is acceptable; 20%+ is strong.

- Niche / premium: 8-15% RPR at 90 days is acceptable; 15%+ is strong.

If RPR falls below the lower bound for your category, investigate acquisition quality, product-market fit, and onboarding messaging before scaling spend.

Cohort Decay Curves - Reading the Slope

A cohort decay curve plots RPR over time. Day 30 RPR, Day 60 RPR, Day 90 RPR, and Day 365 RPR reveal whether customers are returning faster or slower than expected.

Healthy decay pattern: RPR accelerates in weeks 2-4, plateaus by week 12, then decays slowly. Example: 5% at day 30, 15% at day 60, 25% at day 90, 28% at day 365.

Red flag patterns:

- Flat line: RPR stays at 2-3% across all windows. Indicates product dissatisfaction or poor unboxing experience.

- Cliff drop: RPR jumps to 20% by day 30, then falls to 22% by day 90. Suggests one-time event (gift, promotion) rather than habit formation.

- Inverted slope: RPR lower at day 30 than day 60. Indicates delayed satisfaction or seasonal purchase cycles (rare in DTC).

Set a minimum acceptable slope: RPR should increase by at least 3-5 percentage points between day 30 and day 90. If not, product or messaging needs review.

Churn Triggers - When to Act

Churn triggers are thresholds that signal intervention is needed. Define these before data arrives to avoid reactive decision-making.

Trigger 1 - Cohort RPR miss: If a new cohort's 90-day RPR is 5+ percentage points below the previous three-month average, flag for root cause analysis within 48 hours.

Trigger 2 - Decay curve inversion: If day 60 RPR is lower than day 30 RPR for two consecutive cohorts, pause new customer acquisition and audit product quality.

Trigger 3 - Repeat customer AOV collapse: If repeat customers' average order value drops 20%+ versus their first purchase, investigate product assortment, pricing, or discount dependency.

Trigger 4 - Email engagement cliff: If repeat purchase rate among email-engaged customers (opened 2+ emails in 30 days) falls below 18% for consumables, review email content and frequency.

Document these triggers in a shared ops doc. Assign ownership. Set review cadence (weekly for new cohorts, monthly for mature cohorts).

Segmentation - Retention Isn't Uniform

Aggregate retention masks critical failure modes. Segment by acquisition channel, product category, price tier, and geography.

High-priority segments to track:

- Paid social vs. organic: Paid social cohorts often show 30-50% lower RPR than organic. This is normal but must be monitored for deterioration.

- First-time discount vs. full price: Discount-acquired customers typically show 40-60% lower RPR. Set separate benchmarks. If discount cohorts RPR falls below 8% for consumables, reconsider discount strategy.

- Product category: If a brand sells both consumables and apparel, track RPR separately. Consumables may show 35% RPR while apparel shows 10%. Mixing them hides category-specific problems.

- Geographic: International cohorts often show 20-40% lower RPR due to shipping friction. Track separately and don't penalize international acquisition based on domestic benchmarks.

Create a segmentation matrix in your analytics tool. Review monthly. If any segment's RPR drops 3+ points month-over-month, investigate before it spreads.

Failure Mode: The Acquisition-Retention Trade-off

A common failure: scaling acquisition spend while retention declines. The math looks good short-term (more customers) but collapses long-term (fewer repeat buyers).

Warning sign: CAC increases 15%+ while RPR decreases 5%+. This signals acquisition is pulling from lower-quality audiences.

Decision rule: Before increasing acquisition spend, confirm RPR is stable or improving. If RPR is declining, hold acquisition flat and invest in retention (email, loyalty, product improvements) for 30 days. Re-measure. Only resume acquisition scaling if RPR stabilizes.

Track the ratio: Repeat Customer Value (RCV) / CAC. RCV = (Average repeat purchase value × repeat purchase frequency × customer lifespan). If this ratio is below 3:1 and declining, acquisition is outpacing retention economics.

Measurement Setup - Avoiding Data Traps

Retention data is only useful if measured consistently.

Trap 1 - Mixing purchase and subscription: If tracking both one-time purchases and subscriptions, measure RPR separately. Subscriptions inflate retention artificially.

Trap 2 - Including exchanges: An exchange (return + rebuy) shouldn't count as retention. Only count net new purchases.

Trap 3 - Ignoring refunds: If a customer buys, returns, then buys again, count them as retained. But if they buy and refund without a second purchase, don't count them.

Trap 4 - Cohort window drift: Stick to calendar-based cohorts (all customers acquired in January) or fixed-window cohorts (all customers acquired in the past 30 days). Don't mix.

Set up a single source of truth in your analytics platform. Document the exact formula. Share with the team. Audit quarterly.

Benchmarking Against Yourself

External benchmarks are useful for calibration, but internal trends matter more. A brand with 18% RPR at 90 days is strong if it was 15% last quarter, and weak if it was 22% last quarter.

Create a rolling 12-month retention dashboard showing:

- 90-day RPR by cohort month

- 365-day RPR by cohort month (for mature cohorts)

- Repeat customer AOV trend

- Repeat customer email engagement rate

Review monthly. Set a 12-month trend line. If the trend is flat or declining, retention is a priority. If the trend is up 2-3 points per quarter, retention is healthy and can support acquisition scaling.

Questions

FAQ

What's the difference between repeat purchase rate and retention rate?

Repeat purchase rate measures the percentage of customers who buy again within a window (e.g., 90 days). Retention rate is broader and can include email engagement, app usage, or subscription status. For DTC Shopify brands, repeat purchase rate is the most actionable metric and should be the primary retention KPI.

Should we measure retention differently for subscription vs. one-time purchase products?

Yes. Subscription customers have built-in retention (they're charged automatically), so their RPR will be artificially high. Measure subscription churn (cancellation rate) separately from one-time repeat purchase rate. Don't blend them in a single retention metric.

How long should we wait before declaring a cohort's retention 'final'?

90 days is the standard window for initial assessment. By day 90, 70-80% of repeat purchases have occurred. Day 365 RPR is the true lifetime retention, but waiting a year to make decisions is impractical. Use 90-day RPR for operational decisions; track 365-day RPR for annual planning.

If our retention is low, should we cut acquisition or fix retention first?

Fix retention first. Low retention usually signals product, messaging, or onboarding problems. Scaling acquisition into a broken funnel wastes money. Pause acquisition growth, run retention diagnostics (survey customers, audit product quality, test email sequences), and re-measure after 30 days. Only resume acquisition scaling if retention improves.

Want this on your account?

Thirty minutes. Bring the number that keeps you up.

Book a demo