Definition and formula
Incremental margin is the extra margin a marketing activity produced, as distinct from margin it was merely present for. The loyalty programme guide describes it as profit from behaviour the programme actually changed rather than behaviour it simply rewarded, and the site's retention and direct mail guides apply the same idea to mailings and message types.
Margin per customer assigned to the activity − Margin per customer in the held-out group(currency)
- Margin per customer assigned to the activity is the margin on all orders placed during the reading window by the customers assigned to receive it, taken from the business's order records whichever channel the order came through, divided by the number assigned. Everyone assigned counts, including people who ignored the invitation.
- Margin per customer in the held-out group is the same calculation for comparable customers kept away from the activity, ideally chosen at random before it starts.
- Margin is the business's own figure from its accounts, not revenue. The loyalty pilot uses contribution margin and the direct mail guide gross margin; whichever is chosen is stated and applied the same way to both groups.
- The result is an amount of money per assigned customer, before the activity's own costs: rewards, print and postage, message fees, platform and staff time.
One alias, incremental gross profit, is the measure the win-back offer test compares its arms on, in the repeat purchase and loyalty guide's account of it.
How to interpret it: per customer, in total and net of costs
Per assigned customer. The formula gives a figure per person assigned, which lets activities of different sizes be compared and keeps a launch's real reach in view: people who would ignore the activity after launch dilute the figure in the test too.
In total. Multiplied by the number assigned to the activity, it becomes the margin the activity added in the period, and that total is what the guides weigh against cost. The loyalty guide says a programme pays when incremental margin is larger than the redeemed rewards and the running costs together. The guide to measuring direct mail reads a mailing's return as the incremental margin it added less what it cost, segment by segment. The SMS and messaging page for retention keeps a message type while its incremental margin outruns the cost of each send and the subscribers who opt out.
Net of costs. Taking the activity's own costs off, per assigned customer, gives the net figure that the loyalty programme pilot test uses as its primary metric. That net figure is a separate measure, not an alias, even where the pilot's decision rule shortens it to incremental margin after costs. A positive incremental margin with a negative net figure means the activity changed behaviour but cost more than it added.
As a range and over time. The figure is a difference between two groups and carries noise, so the pilot reads its net figure with the interval around the difference. And if the activity only brought purchases forward, the gap narrows after the window closes, which is why the pilot and the direct mail guide read both groups again later.
What moves it
- How much behaviour actually changes. Rewards and messages that land on purchases customers were already going to make add cost and no margin.
- Which segment is assigned. The same activity can add margin in one segment and none in another, which is why the direct mail guide costs each segment separately.
- The margin on the extra orders. Extra sales of discounted or low-margin items add less; a mailing whose extra sales carry a thin margin may cost more than it brings in.
- How many of those assigned take part. Only part of an invited group joins, and the people who do not still sit in the denominator.
- The reading window. A window too short for customers to buy again misses margin; one that ends before the gap from pulled-forward purchases narrows flatters the result.
What good looks like
The site cites no reliable benchmark for incremental margin, and none is offered here. What counts as good depends on what the activity costs and on the business's own margins, so a figure from another business, category or programme describes different economics. Judge it three ways. Against the activity's own costs per assigned customer: the loyalty pilot works out beforehand the incremental margin that would just cover the programme's costs and sizes the test around it. Against zero, with its interval: a range that straddles zero has not shown that the activity adds anything. And against the business's own earlier readings for the same kind of activity and segment, over a similar window and season.
A coffee roaster reads its rewards scheme against a held-out group
Drexyax Coffee, an invented roaster selling subscriptions and bags online, assigned 12,000 eligible customers at random: 6,000 invited to its rewards scheme and 6,000 held out. After 16 weeks, contribution margin per invited customer was £34.20 against £31.70 for the held-out group, an incremental margin of £2.50 per invited customer (its interval stayed above zero), or £15,000 in total. The first dashboard had compared the 1,500 members with everyone else and showed members earning about £26 more margin each; the held-out comparison put the caused part at roughly £10 per member, so most of the members' extra spend would have happened anyway. Rewards redeemed (£10,200), rewards still owed (£900) and running costs (£5,100) came to £2.70 per invited customer, leaving a net figure of minus £0.20. The roaster redesigned the rewards around second subscriptions, the behaviour that moved, and tested again.
Related metrics
IncrementalityConversions caused by marketing. is the general idea of results caused by marketing; incremental margin prices it in margin rather than conversions. The Holdout GroupA randomly chosen part of an audience kept from exposure to measure what exposure caused. supplies the second half of the formula, and without one the figure cannot be measured. Contribution MarginRevenue left after every cost that varies with each sale., revenue minus the costs that rise and fall with each sale, is the margin the loyalty pilot uses on each order. BreakageThe share of issued loyalty rewards that customers never redeem, because points expire, balances are too small to use or members forget., the share of issued rewards that members never redeem, is not a saving to count on: the pilot values rewards still owed in full. Incremental ROASRevenue caused by advertising, measured against a holdout, divided by the ad spend. asks a related question for advertising, as a ratio of caused revenue to spend; incremental margin is an amount of margin, measured before costs. The repeat purchase and loyalty guide sums up the habit the metric enforces: a programme's full cost is set against incremental margin, not incremental revenue.
Common mistakes
- Comparing members with non-members. People who join were different before they joined, so the gap measures who signs up as much as what the scheme does.
- Using revenue. Revenue hides thin margins on the extra orders.
- Dividing by participants only. Customers who ignored the activity belong in the denominator, as they would after a launch.
- Treating it as the net result. It comes before the activity's costs; the decision needs the net figure as well.
- Assuming breakage. Valuing unredeemed rewards below their full cost makes the net figure look better than the programme is.
- Reading one window only. Purchases brought forward look like extra margin until a later reading shows the gap closing.
- Blending segments. An activity that pays in one segment and loses in another can average out to nothing.
