A group operations director pulls up the loyalty dashboard on a Monday morning. Members spent £184,000 last month. The platform vendor's quarterly review slide says the programme is generating six-figure monthly revenue. The CFO asks the obvious question: how much of that would they have spent anyway? The room goes quiet. That silence is the single most common loyalty measurement problem in hospitality, and it is why most operators cannot honestly answer whether their programme is increasing revenue or just relabelling it.
To measure whether loyalty is actually increasing revenue, hospitality groups need three things in place: a pre-launch baseline of repeat rate, AOV and visit frequency for a comparable cohort; a clear separation between total member revenue and incremental revenue (the lift over what those same customers would have spent without the programme); and a monthly cadence where someone actually pulls, cleans and interprets the data. Without all three, the ROI number is a story, not a measurement.
- Total member spend is not loyalty ROI. Incremental lift is.
- If you launched without a baseline, your first job is reconstructing one from historical POS data, not calculating ROI.
- The three numbers worth tracking monthly: repeat visit rate, AOV lift between members and non-members, and 12-month LTV by cohort.
- Benchmarks differ wildly by vertical: a hotel's repeat rate target is not a casual dining group's.
- A dead programme with enrolled members but no engagement cadence cannot be measured at all, because there is nothing driving movement to attribute.
The baseline problem nobody talks about#
Every loyalty ROI article on the internet leads with the same formula: ROI = (net profit from programme - programme costs) / programme costs. The formula is fine. The inputs are the problem. "Net profit from programme" assumes you can isolate what the programme caused, and that requires a baseline.
A baseline is the answer to: what would these customers have done without the programme? Without one, you are comparing member spend to nothing, which always looks impressive, because the most engaged customers self-select into loyalty programmes. They were going to come back anyway. They enrolled because they already liked you.
For a hospitality group launching loyalty, the baseline work happens before the cards go live. Pull 12 months of POS data. Calculate repeat visit rate (percentage of customers who return within 90 days), average order value, and visit frequency. Segment by site and by daypart if you can. That is your benchmark. Once the programme is live, you compare enrolled members against a matched cohort of non-members with similar pre-enrolment spending patterns. The delta is your lift.
Operators who launched without doing this can still reconstruct a baseline retrospectively, but it takes effort. Eagle Eye's analysis of loyalty ROI argues the same point: without a control group or a historical baseline, the maths is decorative.
Incremental revenue vs total member revenue#
This is the distinction that kills most loyalty business cases when a CFO interrogates them. Total member revenue includes every pound a loyalty member spent, including the pounds they would have spent without the programme. Incremental revenue is only the lift caused by the programme: extra visits, higher AOV, longer customer lifespan.
| Metric | What it tells you | What it does not tell you |
|---|---|---|
| Total member spend | How much enrolled customers spent in your venues | Whether the programme caused any of it |
| Member share of revenue | How concentrated revenue is in known customers | Whether loyalty drove the concentration or just labelled it |
| Incremental revenue (lift) | Extra spend caused by the programme vs a matched non-member cohort | Nothing else, this is the number |
| Repeat visit rate change | Whether members come back more often than they used to | Whether AOV is moving in the same direction |
| 12-month LTV by enrolment cohort | Whether the programme is compounding value over time | Short-term campaign performance |
"Members would have spent some of that money anyway. The lift is what matters. Everything else is a vanity slide."
The three numbers worth tracking every month#
Loyalty measurement does not need a 40-metric dashboard. It needs three numbers tracked consistently.
Repeat visit rate. Of customers who visited in month one, what percentage returned within 90 days? Track members vs non-members. For multi-site casual dining, a healthy member repeat rate sits noticeably above the non-member baseline, often double, depending on cuisine and price point. For boutique hotels, the cycle is annual, so the equivalent measure is repeat booking rate within 18 months.
AOV lift. Average order value for members vs non-members, matched for daypart and site. If members are not spending more per visit than comparable non-members, the programme is driving frequency only. That is fine, but you need to know.
12-month LTV by cohort. Group members by the month they enrolled. Track total contribution margin over 12 months. Cohorts should trend upward as the programme matures. If they flatten, engagement has died.
The dead programme problem#
There is a category of loyalty programme that cannot be measured at all: the one with members enrolled but no active engagement cadence. No campaigns going out. No reward changes. No reason for a customer to think about the card sitting in their wallet. If nothing is happening, there is no signal to measure, and any "ROI" calculation is meaningless because nothing is being driven.
This is the most common state of hospitality loyalty programmes after the first 90 days. Operators buy software, launch with a push, sign up a few thousand members, and then the marketing team gets pulled to a venue opening or a Christmas campaign and the programme goes quiet. Members stop checking. Stamps stop being added. Six months later someone asks for an ROI number and there is nothing to report.
Measurement only has value when something is actually being run. A monthly campaign cadence, periodic reward refreshes, scheduled pass updates, all of these create the signal that measurement captures. If you are looking at a flat line in your dashboard, the answer is not better measurement. It is operating the programme.
Why the data sits unused#
Most platform dashboards show member counts, points issued, redemptions and total member spend. They rarely show incremental lift against a non-member control, because that calculation requires joining loyalty data to POS data and segmenting cohorts. That is a data analyst's job, not a marketing manager's, and most hospitality groups under 50 sites do not have a dedicated analyst.
Which is why the dashboard sits open for ten minutes a week, the operations director notes member count is up, and nobody can answer the CFO's question. The measurement that matters needs someone pulling exports monthly, comparing cohorts, writing a one-page summary. Propello's ROI guide makes the same point: ROI measurement is a discipline, not a feature.
"We build the programme. You run the business."
This is the gap Carrott fills for hospitality groups. We run managed loyalty programmes for hospitality groups end to end: setting the baseline before launch, running the monthly campaign cadence that keeps the signal alive, and producing the cohort analysis that separates incremental revenue from total member spend. You see how we increased repeat visit rate for multi-site operators in a one-page monthly report, not a dashboard you have to interpret.
What a credible monthly loyalty report contains#
- Active member count and net new enrolments, by site
- Repeat visit rate: members vs matched non-member cohort, with the delta
- AOV: members vs matched non-member cohort, with the delta
- Incremental revenue estimate (the delta multiplied by member visit volume)
- 12-month LTV trend by enrolment cohort
- Campaign performance: what went out, redemption rate, attributed revenue
- One-line interpretation of the month, and what changes next month
Seven items. One page. Sent on the same day every month. That is the difference between a programme you can defend in a board meeting and one you cannot.
How do I establish a loyalty baseline if my programme already launched without one?
Pull 12 to 24 months of historical POS data from before the launch date. Calculate repeat visit rate, AOV and visit frequency by site. Use this as a retrospective baseline and compare current member behaviour against a matched non-member cohort with similar pre-enrolment patterns. It is less clean than a pre-launch baseline but defensible.
What is a realistic repeat visit rate lift to expect from a hospitality loyalty programme?
It depends on vertical and the previous repeat rate. Casual dining groups with weak existing repeat behaviour often see member repeat rates 1.5x to 2x non-members in the first year. Groups already strong on repeat see smaller percentage lifts but higher absolute revenue impact. Set targets against your own baseline, not a published benchmark.
Is total member revenue ever a useful metric?
Yes, as a tracking number for share of revenue going to known customers, which has value for CRM and forecasting. It is not a measure of programme ROI, and it should never be presented as one.
How often should we recalculate loyalty ROI?
Monthly for operational metrics (repeat rate, AOV, campaign performance), quarterly for a full incremental revenue and LTV review. Annual reviews are too infrequent to catch a programme dying.
Our programme has members but no active campaigns. Can we still measure ROI?
Not meaningfully. With no engagement activity, there is no programme-driven behaviour to attribute. The first step is building an engagement cadence, then measuring the lift it produces. Measurement of a static programme returns noise.