How to tell if a loyalty programme pays: an example review

The programme. A member earns one point for each whole euro of an order. 500 points buy a €5 voucher, taken off an order, so a point is worth one cent and members redeem in whole blocks of 500. Points last 12 months, counting the month they are earned: points earned in October 2024 expired at the end of September 2025. A redemption uses the member’s oldest points first. Customers join at the checkout, and every customer has an account with the shop, so orders placed before a customer joined are on record.

The points account

Year 1 runs from October 2024 to September 2025, year 2 from October 2025 to September 2026. Points first expired at the end of September 2025, when the programme’s first month of points reached the end of their life, so almost all expiry falls in year 2.

Breakage and the liability

Breakage is the share of points never redeemed. It is measured on points whose life has ended: every point earned in year 1 has now been redeemed or has expired, and 40.0% of them expired. Expiry falls mostly on members who have never redeemed a voucher: 85.4% of all the points that expired belonged to them.

At the end of September 2026, 1,418,012 points are on account. At a cent each, they cost €14,180 if every one is redeemed. They will not all be: if they break at year 1’s rate, 505,334 of them will be redeemed, €5,053. For a liability that is the cautious reading, since members who joined later redeem more slowly, as the next table shows.

How fast members redeem

Each row is the members who joined in one quarter; each figure is the share of the points they earned in their first 3, 6 or 12 months, the joining month counted, that they redeemed in the same months.

Within 3 months members redeem little, at most 5.4% of their points: 98.6% of members earn fewer than 500 points in their first 3 months. By 12 months the members who joined in the first quarter had redeemed 45.4%; each quarter after it redeemed less, from 38.4% to 42.2%.

Do members spend more because of the programme?

Over the last 12 months, members who joined in year 1 spent €32.89 a month and customers who have not joined €6.97, counting everyone who ordered at least once in those months: a gap of €25.92. Most of that gap is who joins, not what joining does. Customers join at the checkout, so those who order most meet the offer most often.

Two comparisons take who joins out of it. The first sets members against themselves: the 4,457 members who joined between April 2025 and March 2026 spent €20.07 a month in the 6 months before joining and €22.23 in the 6 months after, €2.16 more. The second sets each of them beside a customer who had not joined, ordered in the same month and spent nearly the same before: the matched customers spent €19.92 a month before and €19.22 after. Against them, members spent €2.86 a month more, between €2.01 and €3.71 at 95%: 14.9% on top of what their matches spent. The joining month is left out of both, since it holds the order that brought the member in.

The two reads agree: the members’ own change sits inside the matched range. So of the €25.92 gap, €2.86 a month is the programme’s doing, 11.0%; the rest is who joins. At the shop’s 40% gross margin, €2.86 of spending earns €1.14 a member a month. Against it, the same members earned 21.9 points a month, and at year 1’s breakage those points cost €0.13. The programme earns more than 8 times what its points cost, and more than they cost even at the low end of the range, before what the programme costs to run.

Decision: should points last 24 months instead of 12?

The question. Points last 12 months. Letting them last 24 would keep points for the members who now lose them, at the cost of more vouchers and a larger liability.

The answer. Yes. Replayed on the two years, the longer life costs €1,195 a year in extra vouchers and adds €4,662 to the liability. It pays for itself if the members who now lose points spend 0.4% more, which the data cannot show either way.

What the data shows. In year 2, 5,469 members lost 610,388 points to expiry, €6,104 of vouchers. 1,944 of them, 35.5%, have ordered in the last 3 months. Most of them earn slowly: 500 points within 12 months is 41.7 points a month, about €42 of orders, and at a steady rate below that, 12 months of points never make a voucher. 5,010 of the 5,469 earn less than that. Of all members a year or more in the programme, 78.4% earn under 41.7 points a month; at 24 months the line halves to 20.8, and 65.0% earn under it.

What the change does, replayed. We replayed the two years under each rule on two assumptions: members order exactly as they did, and a member who orders holding 500 points or more redeems on 40.4% of such months, the rate in the data. Replayed at 12 months, members redeem 0.9% more points in year 2 than they actually did, close enough to compare the two rules on.

At 24 months, members redeem 119,500 more points in year 2: the change costs €1,195 a year in vouchers. Expiry all but stops, and the points on account rise by 466,247, €4,662 more in the liability. Breakage falls: in the replay, 38.2% of year 1’s points go unredeemed at 12 months (40.0% in fact), and at most 31.9% at 24, where most of the unredeemed ones are still on account with months to run. Had members redeemed on twice as many of those months, 80.7%, the change would cost €1,320 a year, 10% more, and pay for itself at a 0.4% rise in those members’ spending: the answer does not turn on that assumption.

What the data cannot show. Whether members who keep their points buy more. Every member in the history was on the same rule, so nothing in it shows what keeping points does. At a 40% margin, €1,195 a year of vouchers is paid back by €2,988 of extra orders: 0.4% of what the members who lost points in year 2 spent that year. For scale, joining the programme lifted members’ spending by 14.9%, in the comparison above.

Recommendation. We recommend letting points last 24 months. The cost is known, within the replay’s assumptions; the gain rests on members who keep their points spending at least 0.4% more. The recommendation stops holding if keeping points changes nothing in how members buy: the change is then €1,195 a year with no return in spending, and the liability is €4,662 larger.

Not covered

• How the points are treated in the shop’s accounts, which is its accountant’s call.
• What the programme costs to run: its software, staff and marketing.
• The notice an expiry change needs, and whether it applies to points already earned.
• Other changes to the programme, such as the earn rate, the size of a block or rewards other than vouchers.

This review is an example on an invented shop and is never signed, so its stamp stays at “Sign-off pending”.

What it reads. Your members with the dates they joined; orders for every customer, member or not, so spending before joining is on record; and the points ledger, with points earned, redeemed and expired by member and month. From exported files or your warehouse. Where the data goes is set out on data handling.

What you receive. A review like this one: the points account, breakage and the liability, how fast members redeem, what the programme adds to spending, and a decision memo on the change you are weighing, with a method note listing the checks that ran.

How it is checked. Agents build the analysis and draft the review. Every deliverable gets automated number checks and a review by an agent; on what you act on, such as the decision memo, that review is a panel of three reviewer agents and a judge, and each check can send the work back to be fixed; two fix rounds that fail the same way stop the work, and the director decides. The analytics director signs what you act on. The steps are on how it works.

Price. A programme review is a sprint, from €12k, fixed in a one-page proposal before work starts. What a customer is worth over their life, members included, is on customer lifetime value.

To set how long points last and how many a euro earns, and see breakage, the liability and the points’ cost move, run the points demo on this review’s shop.