- Member counts and points issued do not prove a loyalty program works. The question is whether members buy more than they would have without it.
- Track engagement (active members, redemption), behavior (repeat rate, frequency, spend) and cost (rewards, discounts, software) together.
- Compare members with similar non-members, or use a holdout, because loyal customers join programs anyway.
Why is loyalty program ROI hard to measure?
A loyalty program almost always looks successful at first glance. Members spend more than non-members, order more often and stay longer. But that comparison is misleading. Your best customers were already your best customers before they joined. They are also the most likely to sign up.
The real question is: how much extra behavior does the program create, and does that extra behavior cover the cost of points, perks and software? Without answering that, a program can quietly become an expensive discount for people who would have bought anyway.
Which metrics should you track?
Group them into four areas.
1. Participation
| Metric | What it shows |
|---|---|
| Enrollment rate | Share of customers who join |
| Active member rate | Share of members who earned or redeemed in the last 90 or 180 days |
| Members by tier | Distribution across tiers if you have them |
2. Engagement with rewards
| Metric | What it shows |
|---|---|
| Points issued and redeemed | Program activity |
| Redemption rate | Share of points redeemed, a sign rewards feel valuable |
| Time to first redemption | How quickly members experience a benefit |
| Breakage | Points that expire unused, which lowers cost but can signal weak rewards |
3. Customer behavior
| Metric | What it shows |
|---|---|
| Repeat purchase rate | Share of members who buy again |
| Purchase frequency | Orders per member per year |
| Average order value | Spend per order |
| Retention or churn | Share of members still active after 6 or 12 months |
| Customer lifetime value | Total contribution over time |
4. Cost
- Value of rewards redeemed (at cost, not retail price).
- Discounts and free shipping given to members.
- Software and operations costs.
- Marketing time spent on the program.
How do you measure incremental impact?
The core comparison is members versus similar non-members. Ways to make it fair:
Matched comparison
Compare members with non-members who looked similar before joining: same order count, spend level, recency and acquisition channel. Then compare their behavior after the join date. Cohort analysis helps here.
Before and after for the same customers
Compare each member’s purchase rate in the 6 months before joining with the 6 months after. This removes some selection bias, but it is affected by seasonality and natural changes in behavior.
Holdout test
For new launches or major changes, keep a random group of eligible customers out of the program (or out of a specific perk) for a period. This is the most reliable method. See incrementality testing.
How do you calculate program ROI?
A simple framework:
Program ROI = (incremental contribution from members − program costs) ÷ program costs
Worked example (illustrative numbers): a home goods store’s program has 12,000 active members.
- Matched comparison shows members place 0.4 more orders per year than similar non-members.
- Average contribution per order (after product, shipping and fees) is $28.
- Incremental contribution: 12,000 × 0.4 × $28 = $134,400.
- Program costs: rewards redeemed at cost $48,000, software $9,600, member-only free shipping $21,000. Total $78,600.
- ROI = ($134,400 − $78,600) ÷ $78,600 ≈ 71%.
The program pays for itself. But the store notices that 35% of reward cost goes to members who already bought monthly before joining. It shifts some rewards toward experiential perks and early access, which cost less and are valued by its best customers.
What warning signs should you look for?
- Low active member rate: many sign-ups but few earn or redeem. The program may be unclear or rewards too distant.
- High breakage with low engagement: points expire because rewards are not attractive.
- Redemptions concentrated in discounts: the program may be training customers to wait for rewards.
- No difference versus matched non-members: the program is not changing behavior.
- Rising costs without rising frequency: the program is becoming a margin leak.
How should you report on the program?
Monthly, show:
- Active members and enrollment rate.
- Redemption rate and reward cost.
- Repeat purchase rate and frequency for members versus matched non-members.
- Estimated incremental contribution and ROI.
Quarterly, review tier thresholds, reward value and the mix of discount versus non-discount perks. For program design, see the eCommerce loyalty program guide and VIP customers.
Checklist: loyalty program measurement
- [ ] Active member rate tracked, not just total sign-ups
- [ ] Redemption rate, breakage and time to first redemption reported
- [ ] Member behavior compared with matched non-members
- [ ] Before and after analysis for new members
- [ ] Holdout test used for launches or major changes
- [ ] All program costs counted at cost, including free shipping and discounts
- [ ] Incremental contribution and ROI calculated at least quarterly
- [ ] Reward mix reviewed for discount dependence
- [ ] Tier thresholds reviewed against customer value distribution
- [ ] Results shared monthly with marketing and finance
FAQ
What is a good loyalty program redemption rate?
There is no universal benchmark, because it depends on reward design. A very low rate suggests rewards are unattractive or hard to reach. A very high rate may mean rewards are too generous. Watch the trend and its effect on margin.
Why do loyalty members always look better than non-members?
Because your most loyal customers are the most likely to join. That is selection bias. Compare members with similar non-members or use a holdout group to see what the program actually changes.
How long before a loyalty program shows results?
Many programs need 6 to 12 months to show a clear change in repeat purchase behavior, depending on how often customers usually buy. Track engagement metrics in the meantime.