Stop Guessing on Retention: Definitions, Thresholds, and Failure Modes

Three Metrics You Need

Most DTC operators conflate retention with repeat rate, then wonder why their numbers don't match. Three distinct metrics clarify the picture.

Repeat Purchase Rate (RPR) is total repeat customers divided by total customers in a cohort. A brand with 1,000 customers and 250 repeat purchases has a 25% RPR. This is backward - looking. It tells you what happened, not whether acquisition is sustainable.

Cohort Retention Rate measures the percentage of a cohort (customers acquired in a specific week or month) who return within 90 days. If 1,000 customers were acquired in Week 1, and 180 of them purchased again by Day 90, retention is 18%. This is forward - looking and cohort - specific.

Churn Velocity is the rate at which a cohort stops purchasing. If a cohort loses 5% of repeat customers per week after their second purchase, velocity is 5% weekly. Velocity predicts lifetime value and flags acquisition quality issues early.

Decision Thresholds by Category

Thresholds vary by category, price point, and replenishment cycle. Use these as starting points, not gospel.

Consumables (coffee, supplements, skincare) should target 25% - 35% 90 - day retention. Failure below 15% indicates product - market fit or positioning problems. Velocity should not exceed 8% weekly post - second purchase.

Apparel and accessories typically see 10% - 18% retention. Below 8% signals either poor fit prediction or weak brand loyalty. These categories rely more on seasonal repurchase and gifting.

Niche / premium goods (jewelry, high - end home) often see 5% - 12% retention due to longer repurchase cycles. Measure 180 - day retention instead of 90 - day. Velocity above 10% weekly is acceptable here.

Subscription - adjacent (meal kits, beauty boxes) should hit 35% - 50% retention. Below 20% indicates churn in the core product experience, not acquisition quality.

Consumables: 25% - 35% (90d) | Velocity < 8% weekly
Apparel: 10% - 18% (90d) | Velocity < 10% weekly
Premium: 5% - 12% (180d) | Velocity < 10% weekly
Subscription: 35% - 50% (90d) | Velocity < 6% weekly

Five Failure Modes and Diagnostics

Low retention clusters into five patterns. Each has a diagnostic procedure.

Mode 1: Acquisition Quality. Cohorts acquired via paid social or affiliate show 8% retention; organic shows 22%. Procedure: Segment retention by channel. If paid underperforms by >5 points, audit creative messaging and audience targeting. Check if paid audiences are cold (lookalike) vs. warm (video viewers, website visitors). Paid audiences should match organic intent.

Mode 2: Product - Market Fit. Retention is flat across all cohorts at 12%, but customer satisfaction (NPS, review sentiment) is 65+. Procedure: Survey non - repeaters within 14 days of first purchase. Ask 'Would you buy again?' and 'Why or why not?' If >40% say 'No' or 'Maybe,' product fit is the issue, not marketing.

Mode 3: Pricing Friction. Retention drops 8 points when price increases 15%. Procedure: A / B test price in acquisition messaging and at checkout. Measure both conversion and 90 - day retention for each price. If retention falls but AOV rises, pricing is a trade - off; if both fall, pricing is a friction point.

Mode 4: Email Decay. Retention is 18% overall, but customers who received 3+ emails in Week 1 show 24% retention; those who received 0 - 1 show 10%. Procedure: Audit post - purchase email sequence. Ensure first email lands within 2 hours, second within 24 hours, third within 72 hours. Test 3 - email vs. 5 - email sequences. Measure opens and clicks, not just sends.

Mode 5: Fulfillment Delays. Cohorts with 3 - day shipping show 20% retention; 7 - day show 14%. Procedure: Correlate shipping speed with retention by cohort. If the gap exceeds 4 points, logistics is a retention lever. Test expedited shipping as a post - purchase upsell or included benefit for repeat customers.

Cohort Tracking Setup

Retention measurement requires cohort isolation. Most Shopify stores lack this infrastructure.

Step 1: Tag customers at acquisition. Use Shopify's 'Source' field or a custom metafield to record acquisition channel, date, and cohort week. Ensure all customers are tagged within 1 hour of first purchase.

Step 2: Define the 90 - day window. For a customer acquired on January 1, the window closes on March 31. Any purchase between January 2 and March 31 counts as a repeat.

Step 3: Export cohort data weekly. Pull customer acquisition date, first purchase date, repeat purchase date (if any), and repeat purchase value. Use Shopify's API or a third - party tool (Klaviyo, Segment, custom SQL).

Step 4: Calculate retention by cohort. Divide repeat customers by total customers acquired in that cohort. Track this metric weekly and plot it on a dashboard. Retention should stabilize by Week 12 of a cohort's life.

Step 5: Segment by channel, product, geography. Repeat the calculation for each segment. Identify which segments drive the highest retention and which are outliers.

Common Calculation Errors

Retention math is simple, but execution errors are common.

Error 1: Including existing customers. If a brand acquires 1,000 new customers and 500 existing customers purchase again, do not count the 500 in the retention denominator. Retention applies only to new customer cohorts.

Error 2: Measuring repeat rate instead of retention. Repeat rate (total repeat transactions / total customers) conflates new and existing customers. It is useful for LTV modeling but not for diagnosing acquisition quality.

Error 3: Cohorts too small. A cohort of 50 customers is too small; variance will mask signal. Use weekly or bi - weekly cohorts with at least 200 customers each.

Error 4: Ignoring the window. If the window is 90 days but a customer purchases on Day 91, do not count them. Consistency matters more than inclusivity.

Error 5: Not accounting for seasonality. A cohort acquired in November will have different retention than one acquired in June. Track cohorts across full years and compare year - over - year.

Retention vs. Lifetime Value

Retention is a leading indicator of LTV, not a substitute. A brand with 20% retention and $50 repeat AOV has a different LTV than a brand with 20% retention and $150 repeat AOV.

Use retention to diagnose acquisition quality and product fit. Use LTV to set customer acquisition cost (CAC) limits. The rule: CAC should not exceed 30% of Year 1 LTV for paid channels.

If retention is stable but LTV is declining, the issue is repeat AOV or repeat frequency, not retention rate. Audit product bundling, upsell sequences, and pricing strategy.

If retention is rising but LTV is flat, repeat customers are buying lower - value products. Investigate whether repeat customers are cannibalizing higher - margin products or whether the product mix has shifted.

Retention Improvement Priorities

Not all retention levers are equal. Prioritize by impact and effort.

High impact, low effort: Post - purchase email sequences (3 - 5 emails over 30 days). Expected lift: 3 - 8 points. Timeline: 2 weeks to deploy.

High impact, medium effort: Product bundling or subscription options. Expected lift: 5 - 15 points. Timeline: 4 - 8 weeks (product development + testing).

Medium impact, low effort: SMS or push notifications for repeat prompts. Expected lift: 2 - 5 points. Timeline: 1 week.

Medium impact, high effort: Loyalty program or tiered rewards. Expected lift: 4 - 10 points. Timeline: 8 - 12 weeks (design, build, launch).

Start with email sequences. Measure the lift. Then layer in SMS and product bundling. Avoid loyalty programs until retention is above category threshold.

Questions

FAQ

What is a 'good' retention rate?

It depends on category and replenishment cycle. Consumables should target 25% - 35% (90d). Apparel should target 10% - 18%. Premium goods should target 5% - 12% (180d). Subscription - adjacent should target 35% - 50%. If retention is below 50% of the category benchmark, investigate product fit and acquisition quality.

Should retention be measured at 30, 60, or 90 days?

90 days is the standard for most DTC categories because it captures one full replenishment cycle for consumables and allows seasonal trends to emerge for apparel. Use 30 - day retention for high - frequency categories (daily consumables) and 180 - day for low - frequency categories (premium goods). Consistency matters more than the window chosen.

How do I know if low retention is a product problem or an acquisition problem?

Segment retention by acquisition channel. If organic shows 22% retention and paid shows 8%, acquisition is the issue (cold audiences, misaligned messaging). If all channels show 12% retention, product fit is likely the issue. Survey non - repeaters within 14 days of first purchase to confirm. If >40% say they would not buy again, product is the problem.

What is the relationship between retention and customer lifetime value?

Retention is a leading indicator of LTV. A cohort with 20% 90 - day retention will have higher LTV than a cohort with 10% retention, all else equal. However, LTV also depends on repeat AOV and repeat frequency. Use retention to diagnose acquisition quality and product fit. Use LTV to set CAC limits (CAC should not exceed 30% of Year 1 LTV for paid channels).

Want this on your account?

Thirty minutes. Bring the number that keeps you up.

Book a demo