Coupon effectiveness analysis

Coupon effectiveness analysis measures whether a discount code generates enough incremental profit to justify the margin it gives away. A coupon is only effective when it brings in sales that would not have happened without it. If a code is redeemed on 1,000 orders but 700 of those customers would have bought anyway, the coupon only earned 300 incremental orders while discounting all 1,000. The metric is built on incrementality, not on redemption count.

The analysis matters because the most-redeemed codes are often the least effective. A discount that customers find at checkout, after they had already decided to buy, simply hands back margin on sales you were going to make. High redemption looks like success on a dashboard while quietly eroding profit. Effectiveness reframes the question from how many people used the coupon to how many extra sales the coupon actually caused.

The metric is most useful when it captures the full picture of value, not just the first order. A coupon that wins a new customer who returns at full price is more effective than one that gives a one-off discount to a loyal buyer. A good analysis weighs the incremental margin against the discount cost and considers whether the redeemers are new, returning, or already loyal, because the same code can be a smart acquisition tool for one group and pure margin leakage for another.

Effectiveness is incremental gross margin minus discount cost, divided by discount cost, expressed as a percentage return. The accuracy of the result depends entirely on estimating incrementality honestly rather than crediting the coupon for every order it touched.

1. 1

   Incremental revenue

   Estimate the revenue from orders the coupon genuinely caused. Use a holdout group or a baseline conversion rate to subtract the sales that would have happened without the code, rather than crediting every redemption.

2. 2

   Discount cost

   Sum the value of the discount across all redemptions, including the discount given to customers who would have bought anyway. This is the full margin you gave up, not just the cost on incremental orders.

3. 3

   Margin basis

   Work in gross profit, not revenue, where you can. A 20 percent code on a product with a 30 percent margin behaves very differently from the same code on a 70 percent margin product.

4. 4

   Customer segment

   Split redeemers into new, returning, and loyal customers. The same coupon can be effective acquisition for new buyers and pure margin loss for loyal ones, so a blended figure hides the truth.

For example, a 10 pounds off code is redeemed on 1,000 orders, giving away 10,000 pounds in discount. A holdout test shows only 300 of those orders were incremental, worth 18,000 pounds in revenue. At a 40 per cent gross margin those incremental orders earned 7,200 pounds, so effectiveness is 7,200 minus 10,000, divided by 10,000, which is minus 28 per cent. A code that looked like a 1,000-order win actually gave away more margin than it earned back. This is exactly why incrementality and margin, not redemption count, belong at the centre of the calculation.

A metric tree decomposes coupon effectiveness into the drivers that determine whether a discount earns its keep, so a single return figure becomes a map of where value is created or lost. Effectiveness depends on how many extra sales the code drives and how much margin each redemption costs.

The first level splits effectiveness into reach, incrementality, margin impact, and customer mix. Reach covers how many people see and use the code. Incrementality covers what share of redemptions were genuinely caused by the discount. Margin impact covers the discount depth and the product margin it lands on. Customer mix covers whether redeemers are new, returning, or loyal. When effectiveness drops, the tree shows whether the code reached the wrong audience, leaked margin to existing demand, or simply discounted low-margin products.

KPI Tree attaches RACI ownership to each driver, so the accountable owner for discount depth differs from the owner of targeting. When a campaign return falls below the threshold, the change is pushed to the owner of the branch that drove it, rather than ending the campaign with a vague sense that the coupon underperformed.

Effectiveness benchmarks vary widely by margin, product, and goal, so the figures below are guides for reading a result rather than universal targets. A break-even point of zero percent return means the coupon recovered its discount cost in incremental margin and nothing more.

Always read effectiveness against the goal of the campaign. A code aimed purely at first-purchase acquisition can run at a low or even negative immediate return if the new customers it wins go on to buy again at full price. In that case the honest benchmark includes repeat value, not just the margin on the discounted order.

Improving effectiveness means giving the discount only to the customers whose purchase it genuinely changes, on products where the margin can carry it. The biggest gains come from cutting leakage to demand you already had, not from making the offer louder.

The metric tree approach starts by finding which driver is holding effectiveness down. If the incrementality branch is the problem, tighter targeting lifts the return faster than any change to the offer. If the margin branch is the problem, reducing discount depth or excluding low-margin products protects more profit per redemption.

KPI Tree connects each driver to its owner and uses the verified impact loop to confirm whether a change actually improved the return. When a marketer restricts a code to new customers to cut leakage, the loop checks whether the incrementality branch and the overall effectiveness score genuinely rose, so the team learns which levers move the number rather than assuming the tighter rule worked.

1. 1

   Crediting the coupon for every redemption

   Most redemptions include customers who would have bought anyway. Without an incrementality estimate, the analysis overstates effectiveness and rewards margin leakage.

2. 2

   Measuring on revenue instead of margin

   A discount on a low-margin product behaves very differently from the same discount on a high-margin one. Work in gross profit so the analysis reflects the value actually given away.

3. 3

   Ignoring the customer segment

   A blended figure hides that the same code can win new customers and subsidise loyal ones at the same time. Split by segment to see where the coupon helps and where it leaks.

4. 4

   Judging acquisition codes on the first order only

   A code aimed at new customers can look weak on the discounted order yet pay off in repeat purchases. Include lifetime value before retiring an acquisition coupon.