# E-commerce retention reports credit buyers who'd return anyway, Digital Darts CEO says

By MarketScale Newsroom · Published 2026-11-16 · Marketing Tech on MarketScale
Canonical: https://www.marketscale.com/industries/marketing-tech/e-commerce-retention-reports-credit-buyers-whod-return-anyway-digital-darts-ceo-says

> A habitual buyer opens a replenishment email, closes it without a click and reorders on schedule. The report calls it retention.

## Key points

- The repeat rate of a randomly held-out tenth of eligible customers, kept out for one full purchase cycle, is the baseline. Only the gap above it belongs to the loyalty or win-back program.
- A 15% or 20% win-back code sent to a habitual buyer turns full-price revenue into discounted revenue, and the dashboard books that margin loss as a recovered customer.
- Watch for a rising repeat rate alongside a shrinking count of first-time buyers. Together they mean a store is living off its existing base, so put both numbers on the same monthly page.

For two years, a customer has bought coffee beans on a five-week cycle. When week five comes around, he opens a replenishment email, shuts it without a click, and places his order two days later, exactly when he always does. The retailer's email platform books that sale as retention revenue.

Joshua Uebergang, CEO of Digital Darts, uses that scenario in a November 13 column for destinationCRM. He argues that most online retailers can't tell whether their loyalty and win-back programs earned any loyalty at all, because the reports give credit to customers who were coming back anyway. He was responding to an earlier destinationCRM piece by Latane Conant, which argued that a transaction alone no longer proves loyalty.

The cause is ordinary attribution settings. Many email and SMS platforms count a customer as influenced if they clicked a message, opened an email, or in some setups simply had a text delivered within a set window before ordering. Someone who buys on a steady cycle and gets regular messages will often land inside that window, so the platform claims a purchase the customer was always going to make.

## Paying people to keep a habit

Discounts make it worse. Uebergang notes that many coupons go to customers who would have bought without them. Send a habitual buyer a 15 percent win-back code and he'll usually use it, because nobody refuses money. The order was coming at full price. The report shows a recovered customer, and finance quietly takes a margin cut on revenue the business already had.

Points programs follow the same pattern. Points pile up fastest for the most frequent buyers, so redemptions cluster among the people who least need a reason to return. Uebergang calls rewarding them a reasonable courtesy. He just thinks it belongs in the cost of serving good customers, outside the retention column.

Attribution windows, discounts and points all share one flaw: they reward behavior that already existed and then report it as behavior the program created. For a marketing operations lead, that means the line on the dashboard and the effect of the spend are two separate numbers, and only one of them is being measured.

## Why the flattering number wins budget

Uebergang writes that a vanity metric would do no harm if it never left a dashboard, but this one drives budget decisions. Channel reviews put return on spend side by side. Because a retention program can claim revenue from habitual buyers, it usually appears cheaper than the cost of finding a stranger, and money gradually shifts to whichever program produces the most flattering report.

Block Field

Yotpo collected these figures in an October 30 guide by growth marketing manager Ben Salomon, which cautions that some of the studies are several years old and that the multiples vary by industry. They are the reason retention starts most planning conversations as the safe bet. Read next to Uebergang's argument, they also point to a problem. The cost advantage applies to keeping a customer who might otherwise leave. If existing customers already buy 60 to 70 percent of the time, a large share of any repeat rate is baseline behavior the program didn't create.

The harm builds slowly. This year's first-time buyers become next year's best repeat customers, Uebergang writes, so trimming the budget that acquires them leaves a thinner repeat base a year or two down the line. When retention figures later drop, the standard reaction is to spend more on retention. He says the cycle can go on for years before anyone looks at how many truly new customers show up each month.

## Four moves to a smaller, truer number

The fix uses order history the store already has. Uebergang's one condition is that the team accept a smaller figure it can actually defend. His method comes down to four steps:

1. Split customers by how they behaved before the program touched them. Treat as habitual any buyer who placed at least three orders on a regular rhythm prior to launch. Report this group's repeat rate on its own line, and expect it to barely move.
2. Randomly select about one in 10 eligible customers and withhold the win-back flow or loyalty offer from them for a full purchase cycle.
3. Measure everyone else on contribution margin. From each retained order, subtract the discount, the points liability and the messaging cost.
4. Each month, put new-customer count and repeat rate next to each other on one page.

Most of the work falls to the holdout. A program that works should produce lift among one-time and lapsed buyers, the people who actually needed a push. The repeat rate in a random tenth of the list that was held out shows what doing nothing produces, and the program can claim only the gap above that.

> The repeat rate of a randomly held-out tenth of the list is what doing nothing looks like, and only the gap above it belongs to the program.

Marketing teams resist because a holdout looks like leaving revenue on the table. Uebergang puts the cost at one purchase cycle on a tenth of the list. What the team gets back is a number it can defend in a budget meeting. The fourth step matters for the same reason: when the repeat rate climbs while the number of first-time buyers falls, a store is living off its existing base, and putting the two on one page makes that hard to explain away.

## The margin math behind a recovered order

Uebergang's sharpest point concerns discounts. When a 20 percent code recovers an order from someone who would have paid full price, the result is worse than doing nothing. Revenue reporting can't reveal that. Margin reporting reveals it right away.

Yotpo's guide has a worked example of how much the reporting basis changes the picture, applied to lifetime value. A sample store with a $26 acquisition cost and a $300 customer lifetime value has a ratio of about 11.5 to 1. At a 40 percent gross margin, lifetime value falls to $120 and the ratio to about 4.6 to 1. That's still healthy, but it's less than half the headline figure. The guide also cites an Improvado estimate that using revenue instead of margin-adjusted revenue can inflate lifetime value by 30 to 50 percent.

Uebergang's netting list stops at discounts, points and messaging, and he doesn't give a numeric threshold for switching off an offer for a given segment. Teams building the calculation will need to decide which other costs to include.

Block Field

## Aim the program, then fund the hunt

The obvious pushback is that retention genuinely costs less, and the multiples back that up. Uebergang doesn't dispute it. His point is that when a holdout comes in well below the reported figure, that is rarely grounds for cutting the program. The better move is to aim it: direct spending at lapsed and one-time customers, since those are the groups where the holdout reveals a genuine lift, and let regulars go on buying at full price. Yotpo's guide arrives at a similar conclusion from the other direction. It notes that retention has real costs in rewards, service and software, and that new buyers are how a store replaces customers who drift away.

Where the gap shows up depends on the business. Stores whose revenue comes mostly from consumables reordered on a steady cycle could see a small difference between the holdout and the program group. Stores with mostly one-time buyers may find the program earns more of its credit. Either way, the holdout replaces a guess with a measurement.

Uebergang wants any budget that gets freed up to go back into reaching people who have never bought, through search, social or Google Ads campaigns designed for new customers. He would measure that spending by new-customer count and first-order margin instead of blended revenue, the same bar retention is now held to.

He recommends opening with a single question to the person responsible for the retention report: how many of these customers would have returned without us? If nobody has an answer, start with the next win-back send. Hold back a tenth of the list, and once one purchase cycle has passed, compare the two groups.

## Sources

- [E-Commerce Customer Retention Reports Are Grading the Wrong Customers](https://www.destinationcrm.com/Articles/Web-Exclusives/Viewpoints/E-Commerce-Customer-Retention-Reports-Are-Grading-the-Wrong-Customers-177508.aspx)
- [Customer Acquisition vs Retention Cost: Full Guide](https://www.yotpo.com/blog/cost-of-customer-acquisition-vs-retention/)
- [How to Reduce Ecommerce Return Rates in 2026](https://www.clickpost.ai/en-us/blog/reduce-ecommerce-returns-rate)

Tags: Digital Darts, customer retention, loyalty programs, e-commerce, marketing measurement, holdout testing, customer acquisition cost, contribution margin, CRM, marketing budget

---
Source: MarketScale, https://www.marketscale.com/industries/marketing-tech/e-commerce-retention-reports-credit-buyers-whod-return-anyway-digital-darts-ceo-says. Published for AI indexing and citation; cite the canonical URL. Site guide for agents: https://www.marketscale.com/llms.txt
