Discount addiction: how online stores train customers to wait for a sale, and how to stop
The coupon that pays for itself on the dashboard and quietly costs you on the P&L. Here's how to tell the difference, and how to stop.
Key takeaways
- Discount addiction is when customers who would pay full price learn to wait for your next code, so discounts stop creating sales and start relabeling them.
- At a 45% gross margin, a 20% discount has to lift orders by 80% just to break even on margin.
- Field research found that deeper discounts helped bring first-time customers back but reduced later spending by established customers.
- The same coupon can be a good acquisition tool and a bad retention habit, so target discounts by customer, not by calendar.
- Run every promotion with a random holdout, and judge it on margin caused, not revenue attributed.
Discount addiction is what happens when a store discounts so often that customers learn to wait for the next code. The sales keep coming, but more of them are orders that would have happened at full price anyway. So the discount stops creating demand and starts relabeling it, at a lower margin. This guide explains how discount addiction forms, why a promotion can look like a hit while losing money, how to tell whether your store has it, and a step-by-step way to stop.
What is discount addiction?
Every discount does two things at once. It persuades some people to buy who otherwise wouldn't. And it hands a lower price to people who were going to buy anyway. The first group is the point of the promotion. The second group is its cost.
Discount addiction is when the second group keeps growing. Customers notice the rhythm: a code every few weeks, a bigger one at Black Friday, a "we miss you" offer if they go quiet. Reasonable people respond by waiting. Nobody pays full price on Thursday for something that's 20% off on Monday.
The name sounds dramatic, but the mechanics are dull. Nothing about the customer changes except their sense of timing. That's what makes it hard to spot. The dashboard still shows orders, and the email platform still reports revenue. What changes is how many of those orders the discount actually caused.
If your store runs promotions on a predictable calendar, assume some customers have learned the calendar, and measure how many.
Why customers learn to wait for a sale
Researchers have been watching this happen for decades. A study by Mela, Gupta and Lehmann, published in the Journal of Marketing Research in 1997, used eight and a quarter years of shopper panel data for one frequently bought packaged good. Its results were consistent with the idea that shoppers became more sensitive to price and promotions over time, as brands cut advertising and ran more promotions. It's an old study about a supermarket product, not an online store, so treat it as the classic version of the pattern rather than a measurement of yours.
The more recent signal comes from shoppers themselves. In the 2026 RetailNext Shopper Sentiment Report, a February 2026 survey of 1,053 US adults, 55.9% said they preferred straightforward price discounts over other kinds of promotion, against 19.8% for loyalty rewards. Retail Brew's coverage of the report quoted its warning that the more a retailer relies on discount events, the more shoppers put off purchases until those events arrive.
Put those together and you get the core problem. Customers like discounts, and they're good at learning when they'll get one. Your promotional calendar isn't a secret. It's a schedule your best customers can plan around.
Treat a predictable promotion as something customers will adapt to, and expect the adaptation to show up in your full-price weeks first.
Why a 20% discount needs 80% more orders
Here's the part that surprises people. A discount comes straight out of your margin, not your revenue. So the lift a promotion needs to break even is far bigger than the discount itself.
Break-even order lift = gross margin ÷ (gross margin − discount)
At a 45% gross margin, a 20% discount leaves 25 cents of margin on each dollar instead of 45. To make the same total margin, you need 0.45 ÷ 0.25 = 1.8 times as many orders. That's an 80% lift, from a promotion your team may describe as "only 20% off". Most sale emails would be thrilled with half that.
| Gross margin | 10% off | 20% off | 30% off |
|---|---|---|---|
| 60% | 1.20× (+20%) | 1.50× (+50%) | 2.00× (+100%) |
| 45% | 1.29× (+29%) | 1.80× (+80%) | 3.00× (+200%) |
| 35% | 1.40× (+40%) | 2.33× (+133%) | 7.00× (+600%) |
The bottom-right cell is not a typo. At a 35% margin, a 30% discount leaves five cents in the dollar, and you need seven times the orders to stand still.
Before you approve a promotion, look up its break-even lift in this table. If you've never seen a promotion produce that lift, the promotion is a cost, not an investment.
How a promotion can look like a hit and still lose money
Here's a worked example with an illustrative store: 2,167 customers who ordered in the last 12 months, an $86.80 average order and a 45% gross margin. The store emails a 20% sitewide code to its active customers. It holds back a random 10%, 217 customers, who get nothing. The response rates below are assumptions for the example, not benchmarks.
| Customers | Orders in 14 days | Order rate | |
|---|---|---|---|
| Sent the 20% code | 1,950 | 146 | 7.5% |
| Held out, no code | 217 | 13 | 6.0% |
| Sent group's orders expected anyway (6.0% of 1,950) | 117 | ||
| Orders the code caused | 29 |
The email platform reports a winner: 146 orders × $69.44 discounted average = $10,138 in attributed revenue. Now look at margin. Each full-price order keeps $39.06 (45% of $86.80). Each discounted order keeps $69.44 − $47.74 in product cost = $21.70.
| Orders | Margin per order | Total margin | |
|---|---|---|---|
| With the code | 146 | $21.70 | $3,168 |
| Without it (the 117 who'd have bought anyway, at full price) | 117 | $39.06 | $4,570 |
| Margin the promotion cost | −$1,402 |
The campaign that "made" $10,138 actually cost the store $1,402. To break even, it needed about 211 orders, a 10.8% order rate, instead of 146. It's the 20% code sent to someone who was already typing in their card number, multiplied by 117.
A discount's real price is the margin you hand to people who were buying anyway. Attributed revenue counts those people as wins.
Never judge a promotion by attributed revenue. Judge it by margin from the sent group minus the margin the holdout says you'd have made anyway. The incrementality vs attribution guide explains the method in full.
Do discounts hurt your best customers?
This is where the research gets useful, because it shows the same discount behaving differently for different customers.
In 2004, Anderson and Simester published three field experiments with a mail-order catalog firm that sells durable goods. Customers were randomly sent catalogs with deeper or shallower discounts, then tracked through later catalogs. In the authors' words, deeper discounts "increased future purchases by first-time customers" but "reduced future purchases by established customers."
The established-customer result is the one to remember. In that study, established customers who got the deeper discounts spent $584.68 each in later catalogs, against $733.50 for those who didn't, about 20% less. They bought fewer items and cheaper ones. The authors consider several explanations, including forward buying: stocking up at the low price instead of buying later.
A 2006 study by Michael Lewis in the Journal of Marketing Research looked at the other end of the relationship, using customer data from a newspaper and an online grocer. It found that the deeper the discount used to acquire a customer, the lower their repeat-buying rate and long-term value tended to be. That's value lost before the first order even ships. The customer lifetime value guide shows how to measure it on margin, not revenue.
So the honest answer is: it depends who gets the discount. A first-order offer can bring in a customer who sticks. A blanket code to people who already love you can teach them to spend less. One coupon, two very different results. It's the one marketing tool that's both medicine and junk food, depending on who swallows it.
Stop sending the same offer to new and established customers. Decide who each promotion is for before you decide how deep it goes.
How to tell if your store is addicted to discounts
You don't need a model to find the warning signs. Your order export and a spreadsheet will do. Check these over the last 12 months:
| Signal | How to check | What it suggests |
|---|---|---|
| Rising share of orders with a code | Orders with a discount ÷ all orders, by month | Discounts are relabeling more of your normal demand |
| Full-price dips before each sale | Daily orders in the week before promotions vs normal weeks | Customers have learned your calendar |
| Repeat orders cluster on promo days | Share of repeat customers' orders placed during promotions | Your regulars are timing their purchases |
| Customers who have never paid full price | Customers whose every order used a code | A segment that may only buy on promotion |
| Sales emails beat their holdout by less each time | Lift over holdout for each promotion, in order | The extra demand is wearing out |
No single signal proves it. A store that sells winter coats will see orders cluster in November whatever it does. But three or four of these moving the wrong way together is a pattern worth acting on.
The most useful number is the last one. If you've never run a promotion with a holdout, start with the next one, because that's the only way to see how much of the response would have happened anyway.
Run this check once a quarter, and treat a rising share of discounted orders as a cost trend, not a sales trend.
How to stop discounting without wrecking a quarter
Going cold turkey is how stores end up reinstating the sale in week three. Wean gradually, and measure each step.
- Kill the sitewide code first. Blanket discounts give the most margin to the customers who need it least. Replace them with targeted offers before you change anything else.
- Score discount sensitivity per customer. Sort customers into those who mostly buy at full price and those who mostly buy on promotion. The customer segmentation guide covers how to build that "deal-driven" group. Send codes to them, and full-price reminders to everyone else.
- Hold out a random group from every promotion. Even 10% is enough to show whether a large promotion creates demand. The holdout sample size calculator page has the formula for sizing it.
- Step the depth down. If 20% off is the habit, try 15%, then 10%, each with a holdout. Many stores find a smaller offer keeps most of the incremental orders and much of the margin.
- Make the calendar less predictable. A sale every first Friday is a schedule. Fewer, less regular events give customers less reason to wait.
- Give your best customers something other than money off. Early access, a restock alert or a note from the founder rewards loyalty without teaching anyone to wait. Customers who already pay full price aren't asking for a coupon.
- Rethink the welcome offer. Given the acquisition research above, test a smaller first-order discount against your current one. Judge it by the second-order rate and margin of each cohort, not by first-order conversion alone.
The goal isn't zero discounts. Clearing seasonal stock, launching a new product and winning a first order are all fair uses. The goal is that every discount is sent on purpose, to someone it changes.
Pick the first two steps this month, and run your next promotion with a holdout so you have a baseline before you change anything else.
When discounting is the right call
Some situations make discounts sensible, and some make this advice harder to apply.
- Clearing stock. A discount on last season's inventory turns dead stock into cash. The break-even math changes when the alternative is writing the stock off.
- Very small stores. With a few hundred customers, a 10% holdout may be too small to show a reliable difference. Run holdouts across several promotions and read them together.
- Price-led categories. If your store competes on price and your customers know it, they're comparing you with alternatives, not waiting for you. The fix there is pricing, not promotion discipline.
- The research isn't your store. The studies above cover a packaged good, a catalog firm, a newspaper and an online grocer. They show a pattern worth testing, not a number to assume. Your holdout results beat any study.
If your store fits one of these, keep the holdout habit anyway. It's the one part of this guide that applies everywhere.
In LiftSignals
LiftSignals scores every customer's discount sensitivity every night from their order history, so you can send codes to customers who usually buy on promotion and leave full-price buyers out. Each audience syncs to Klaviyo, Meta, Google and other tools with a random holdout, so every promotion shows the orders and margin it caused.
What to do next
You now know what discount addiction costs: the margin handed to customers who were buying anyway, which attributed revenue counts as a win. The fix starts with knowing who those customers are. Discounts and returns scores every customer's discount sensitivity, so your next code goes to the people it changes and your full-price regulars keep paying full price. Connect your store and see your own deal-driven segment during the 30-day free trial. No card needed, and no coupon either.
Sources
- Anderson, Eric T. and Duncan I. Simester, "Long-Run Effects of Promotion Depth on New Versus Established Customers: Three Field Studies," Marketing Science 23(1), 2004. https://www.kellogg.northwestern.edu/faculty/anderson_e/htm/personalpage_files/Papers/Long_Run_Effects_of_Promotion_Depth_on_New_versus_Established_Customers.pdf
- Lewis, Michael, "Customer Acquisition Promotions and Customer Asset Value," Journal of Marketing Research 43(2), 2006. https://doi.org/10.1509/jmkr.43.2.195
- Mela, Carl F., Sunil Gupta and Donald R. Lehmann, "The Long-Term Impact of Promotion and Advertising on Consumer Brand Choice," Journal of Marketing Research 34(2), 1997. https://scholars.duke.edu/display/pub1485948
- "2026 RetailNext Shopper Sentiment Report," RetailNext, Mar 2026. https://retailnext.net/blog/2026-retailnext-shopper-sentiment-report
- Newman, Andrew Adam, "Discount events are 'training' shoppers to never pay full price: report," Retail Brew, Mar 2026. https://www.retailbrew.com/stories/2026/03/24/discount-events-are-training-shoppers-to-never-pay-full-price-report
Discount addiction questions
What is discount addiction in ecommerce?
It's when a store's regular discounting teaches customers to delay purchases until the next sale. The discounts keep producing orders, but a growing share of those orders would have happened anyway at full price, so the store gives away margin for little extra demand.
How do I know if my customers are addicted to discounts?
Look for a rising share of orders that use a code, full-price weeks that get quieter before each sale, and repeat customers whose orders cluster on promotion days. The clearest test is a holdout: if customers who didn't get the code buy almost as much, the discount isn't creating demand.
Can I stop discounting without losing sales?
Usually you lose some discounted orders and keep most of the margin. Cut gradually: make offers smaller, stop sitewide codes first, and keep discounts for customers who only buy on promotion. Measure each step against a holdout so you can see what actually changed.
Do discounts lower customer lifetime value?
They can. A 2006 study of a newspaper and an online grocer found that deeper acquisition discounts were linked to lower repeat-buying rates and lower long-term customer value. A discount also lowers the margin on every order it touches, which is what lifetime value should be built on.
Put it to work: Discounts and returns → · Incrementality testing → · Holdout sample size calculator →
Keep reading
All discounts & pricing guides →Start here · Customer lifetime value
Customer lifetime value: what it is, how to calculate it, and how to raise it
The number that tells you how much a customer is worth, and therefore how much you can afford to spend winning and keeping them.
5 min read · Updated Nov 15, 2026
Start here · Incrementality & measurement
Incrementality vs attribution: what each one tells you, and when to use which
Attribution says who touched the sale. Incrementality says whether the sale would have happened anyway. They answer different questions.
4 min read · Updated Nov 15, 2026
Start here · Segmentation & RFM
Customer segmentation for ecommerce: types, examples and how to start
How to split customers into groups that behave differently, so each group gets the message, offer and budget that fits.
4 min read · Updated Nov 15, 2026
Send the code only to customers who need it
Score every customer's discount sensitivity, keep codes away from full-price buyers, and see the margin each promotion actually caused.