TL;DR Summary
Loyalty programme ROI is notoriously hard to prove. Here's which metrics actually matter, how to set a baseline before launch, and what tools track them.
Why Loyalty Programme ROI Is Hard to Prove — and How to Finally Measure It
Most loyalty programmes run for months before the owner realises they have no idea whether it is working. The measurement problem is structural: without a baseline, you cannot prove that any improvement in visit frequency or spend is caused by the programme rather than seasonal trends or organic growth. Here is how to set up proper measurement before you launch — and which five metrics actually predict long-term ROI.
Why Loyalty ROI Feels Impossible to Prove
The frustration with loyalty programme measurement is a recurring theme in small business discussions. Owners invest in a programme, see some signups, notice that regular customers seem to be visiting more often, and then struggle to translate that impression into a number they can defend.
The measurement problem has three root causes.
First, most loyalty programmes — especially paper-based ones — produce no data at all. You cannot measure what you are not tracking.
Second, businesses rarely set a pre-launch baseline. Without knowing what visit frequency and average spend looked like before the programme launched, any post-launch improvement is unattributable. It could be the programme. It could be the new menu item. It could be seasonal variation.
Third, the metrics that are easiest to measure (total signups, total redemptions) are not the metrics that predict revenue impact. Signup counts tell you nothing about whether the programme is changing behaviour. Redemption counts tell you nothing about whether the programme is profitable.
The 5 Metrics That Actually Matter
These five metrics, tracked together, give a complete picture of loyalty programme performance. They are listed in order of importance.
1. Visit Frequency Delta
What it is: The change in average visits per month for enrolled customers, compared to their pre-enrolment baseline.
Why it matters: This is the core behaviour you are trying to change. If enrolled customers are not visiting more frequently than before enrolment (or more frequently than non-enrolled customers), the programme is not generating incremental revenue.
How to measure: Track the average number of visits per month for each enrolled customer in the 60 days before enrolment, then compare to the 60 days after. A positive delta means the programme is influencing behaviour. A flat or negative delta means it is not.
2. Average Spend Per Visit (Enrolled vs. Non-Enrolled)
What it is: The average transaction value for loyalty programme members compared to non-members, measured over the same time period.
Why it matters: A loyalty programme that increases visit frequency but reduces average spend (because customers are redeeming rewards rather than buying) may have a neutral or negative revenue impact. You need both dimensions.
How to measure: Segment your transaction data by programme membership status and compare average transaction values. A well-designed programme should show enrolled customers spending equal to or more than non-enrolled customers, even accounting for reward redemptions.
3. Churn Rate (Enrolled vs. Non-Enrolled)
What it is: The percentage of customers who stop visiting over a defined period, compared between enrolled and non-enrolled segments.
Why it matters: 70% of first-time restaurant guests never return. The loyalty programme's primary job is to move customers from that 70% into the returning category. Churn rate is the direct measure of whether it is doing that.
A BJJ gym discussed in Reddit small business communities had 48.7% annual churn even with subscription-based memberships — a reminder that churn is not just a retail problem, and that it persists even when customers have already paid. A loyalty programme that does not measurably reduce churn below the baseline is not earning its keep.
How to measure: Define "churned" as no visit in 60 days (adjust for your typical purchase cycle). Track what percentage of enrolled vs. non-enrolled customers cross that threshold each month.
4. Signup Completion Rate
What it is: The percentage of customers offered a loyalty programme who actually enrol.
Why it matters: The most common reason loyalty programmes underperform is that most customers never joined. A programme that 85% of customers declined to join cannot produce population-level behaviour change. The signup rate sets the ceiling on everything else.
Benchmarks: App-based programmes achieve approximately 15% completion. Paper punch cards achieve around 70%. Digital wallet cards (QR code to Apple/Google Wallet) achieve approximately 95%.
If your signup rate is below 50%, fixing the signup mechanism will have more impact on programme ROI than any other single change.
5. Redemption Rate
What it is: The percentage of enrolled customers who earn and redeem at least one reward within a defined period.
Why it matters: Very low redemption rates (under 10%) suggest your reward threshold is too high or your customers are not visiting frequently enough to reach it. Very high redemption rates (over 60%) suggest your reward threshold is too low and the programme may be functioning as a routine discount rather than an earned incentive.
The sweet spot depends on your business, but 20–40% redemption in the first 12 months is typically a healthy signal — it means the programme is driving repeat visits without subsidising customers who would have returned anyway.
How to Set Your Baseline Before Launch
This is the step most businesses skip, and it is the most important.
Before you launch a loyalty programme, spend four weeks measuring these numbers manually:
- Total unique customers per week. Count individual transactions, not total orders.
- Average transactions per customer per month. If you do not have POS data, estimate this from your transaction count divided by your estimated regular customer count.
- Average spend per transaction. Your POS should have this.
- Percentage of customers who return within 30 days. This requires some tracking — either through card payments (which can be matched by last-4 digits over multiple visits) or through a manual tally.
These four numbers form your baseline. Every improvement after programme launch can be measured against them.
Without this baseline, you will spend months guessing. With it, you will have an answer within 60 days of launch.
What Tools Actually Track This
The right tool depends on whether you have a digital loyalty programme. Paper programmes produce none of this data — which is reason enough to switch.
POS-integrated loyalty platforms (Square Loyalty, Toast's loyalty module, Clover Rewards) automatically connect transaction data to customer profiles. The advantage is seamless integration. The disadvantage is that you are locked into a specific POS ecosystem, and if you change POS providers, you typically lose your loyalty data history.
Standalone digital wallet platforms like GPASS track visit frequency, signup completions, and point balances independently of your POS. This is more flexible for businesses that switch POS systems or operate across multiple locations with different setups. The tradeoff is that spend data requires a manual integration or a separate check-in workflow.
CRM tools (Klaviyo, Mailchimp with loyalty integrations) are useful for tracking email-based engagement but miss in-store visit data unless explicitly integrated with a loyalty card or POS system.
Spreadsheet-based tracking is viable for small businesses in the first 60–90 days of a programme, particularly for establishing the baseline. It does not scale, but it is better than no tracking.
The Attribution Problem
Even with perfect data, loyalty programme ROI involves an attribution problem: you cannot know for certain whether a customer would have visited anyway, or whether the programme caused the visit.
The practical solution is to compare two cohorts:
- Customers who enrolled in the programme
- Customers of similar prior visit frequency who were not offered or did not enrol in the programme
If enrolled customers show meaningfully higher visit frequency over 6 months, the programme is almost certainly driving incremental behaviour — even if you cannot attribute individual visits with certainty.
This is why signup rate matters so much. A programme with 15% enrollment cannot produce a statistically meaningful comparison cohort. A programme with 95% enrolment means almost every customer is in the treated group — and the comparison becomes your pre-launch baseline.
A Simple Measurement Framework
If you want to start measuring loyalty programme ROI this month, use this framework:
| Metric | Measure Before Launch | Measure Monthly Post-Launch | Success Signal |
|---|---|---|---|
| Visit frequency | Average visits/month per customer | Delta vs. baseline | +15% or more in 90 days |
| Average spend | Average transaction value | Enrolled vs. non-enrolled | Equal or positive |
| 30-day return rate | % of customers back in 30 days | Enrolled vs. non-enrolled | +10 percentage points |
| Signup rate | N/A | % of customers who enrolled | Above 70% |
| Churn rate | % with no visit in 60 days | Enrolled vs. non-enrolled | Enrolled churn lower |