Incrementality measures the revenue, conversions or customer value a marketing activity generates beyond what would have happened anyway. It separates genuinely additional outcomes from those a business would have captured regardless, using holdout groups or controlled experiments rather than last-click attribution.
Incrementality is the difference between the outcomes a campaign produced and the outcomes that would have occurred without it. For promotions teams it answers the question discount reporting cannot: did this offer create new revenue, or did it hand a discount to customers who were already going to buy at full price?
Promotional metrics flatter performance easily. A code with a high redemption rate looks successful, but if a large share of redeemers were existing customers already midway through checkout, the promotion gave away margin and added nothing. Incrementality strips out that baseline demand and reports only what the campaign added on top.
The main method is comparison against a control. You measure what happened in a group exposed to the promotion, compare it with a similar group that was not exposed and treat the gap as the incremental effect. Other methods refine this principle.
Common approaches include:
Whichever method a team uses, the principle remains the same: the promotion receives credit only for outcomes above the counterfactual, not for every transaction that carried its code.
Promotions carry a direct, visible cost. Every redeemed code has a margin cost attached, so whether that cost bought new revenue or subsidised existing demand is a commercial question, not an academic one.
Non-incremental spend has a few recognisable causes. Generic codes leak to aggregator sites, where a customer who arrived ready to pay full price finds one and applies it, which is the problem Uniqodo's Code Distribution product addresses by issuing a unique code per publisher rather than one shareable string. Blanket offers reach loyal customers who need no incentive at all. And coupon stacking deepens the discount on an order without adding a single unit of demand. Every one of these shows up in a redemption report as a success.
The same discount spend can represent very different shares of the revenue it sits against. Across 90 merchants and 4,124 promotions in H1 2026, Uniqodo platform data shows referral discounts represented 11.3% of the basket value they generated, against 7.3% for standard promotions. A mechanic can carry a smaller absolute discount per order and still cost more as a proportion of the revenue it produces. Incrementality estimates how much of that revenue was genuinely added. The cost figure tells you what the addition cost. A redemption report gives you neither.
Measurement finds the waste. Controls are what stop it recurring. Single-use codes, qualification rules and eligibility conditions in Uniqodo's Promotion Engine let teams withhold a discount from segments that convert reliably at full price, so the offer reaches the customers whose behaviour it might actually change.
Attribution and incrementality answer different questions. Conflating them can lead to misleading e-commerce measurement.
Attribution asks which touchpoint should receive credit for a conversion. Incrementality asks whether the conversion would have happened at all. A last-click model can assign 100% of a sale to a discount code. That assignment may be accurate for attribution yet meaningless for incrementality because the customer might have converted without the code.
As a result, attribution data can overstate promotional performance. Codes often appear near the end of the purchase journey, so they may collect credit for demand created earlier by brand, product and price. Teams that plan promotional budgets using attributed revenue alone risk over-investing in offers that capture existing demand while under-investing in mechanics that create additional demand.
The two methods work together. Attribution shows where conversions flow and which partners or channels should receive credit. Incrementality testing assesses whether the activity behind those conversions is worth funding.
You do not need a data science team to begin. Start with the promotions carrying the largest discount cost and ask the counterfactual question of each: who is redeeming this, and what would they have done without it? Even a simple holdout on one campaign can provide more useful evidence than a long series of redemption reports.
Next, build incrementality checks into promotion design instead of reviewing performance only after a campaign ends. Restrict codes to their intended audiences, exclude segments that convert reliably at full price and test one variable at a time. The aim is not necessarily to run fewer promotions. It is to assess whether the margin given away generates revenue that the business would not otherwise have captured, and to use that evidence when planning future promotional strategy.
No. Redemption rate measures how many customers used the code, not how many bought because of it. A promotion can redeem heavily and still be close to zero incremental if most redeemers were already at checkout. The two numbers are unrelated, which is why redemption reporting cannot answer the incrementality question on its own.
Yes, with less certainty. Geo testing uses matched markets as the control, matched audience analysis compares statistically similar customer groups after the fact, and baseline modelling forecasts what sales would have been without the promotion. All three are weaker than a randomised holdout, and all three beat assuming that attributed revenue is incremental revenue.
Incremental ROI divides the additional revenue a promotion generated by the cost of running it, counting only revenue above the counterfactual. Standard promotional ROI counts every order that carried the code, so it will always return the higher figure. The gap between the two numbers is the demand the business would have captured anyway.
Stop code leakage. Replace shareable generic codes with high-entropy unique strings. Protect your margins by ensuring discounts only apply to the intended audience under specific, validated conditions.
Execute complex campaigns. Move beyond basic discounts with multi-tiered rewards, product bundles, and discounts, all managed without waiting for a developer to clear your roadmap.
Convert with intent. Use real-time data to trigger onsite nudges or referral loops exactly when they matter. Create a unified journey that turns browsing interest into confirmed sales.
Scale partner sales. Automate the delivery of unique codes to thousands of partners instantly. Replace manual spreadsheets and CSV exports with secure, trackable API distribution.
We'll show you exactly how Uniqodo handles your use case - fraud controls, mechanic complexity, and ROI attribution included.