On a Tuesday afternoon in October, a marketing manager at a six-site restaurant group opens a spreadsheet she has been avoiding. The loyalty programme launched in March. The software bill is £180 a month, which is fine. What is not fine is the column next to it: the hours she has personally spent on the programme since launch. Brief campaigns. Write member emails. Chase the designer. Pull reports the founder asked for. Segment a list she never quite finishes segmenting. She adds it up. It is more than a day a week. The programme has 4,200 enrolled members. She has no idea how many of them have come back.
This is the cost nobody puts in the proposal. A loyalty programme costs whatever the software charges, plus the fully-loaded salary of the person running it, plus the opportunity cost of everything that person is not doing instead. For most hospitality operators, that second number is three to five times the first. And when nobody has the capacity to actually run it, the programme produces enrolment numbers, not repeat revenue.
- Loyalty software is typically 15 to 25% of the true cost of running a programme. Staff time is the rest.
- Most business cases underestimate operations and marketing costs by 30 to 50%, according to CFO-ready loyalty modelling.
- A programme with no operator produces passive enrolment, not active members, and 83% of loyalty apps are uninstalled within 30 days.
- Active member rate is the metric that correlates with repeat revenue. Enrolment is a vanity number.
- Done-for-you loyalty exists because the execution gap is structural, not a skills problem.
The loyalty programme your team launched but nobody runs#
Every hospitality operator we talk to has a version of the same story. They bought a loyalty platform. They printed the cards or set up the wallet pass. They put the QR code on the menu. The founder told the team it would drive repeat visits. And then everyone went back to their actual jobs, because the marketing manager already has a full week and the operations director is solving a different fire.
Six months in, the dashboard shows a four-figure enrolment number, which looks like progress. But nobody has segmented the lapsed members. Nobody has written a win-back. Nobody has built a campaign around the slow Tuesday lunch slot. The programme is live. It is not running.
"Ambition outpaces resources. Operational complexity slows delivery. Most loyalty teams cannot execute the programmes they have designed."
The Open Loyalty 2026 trends report frames this as a structural issue, not a competence one. The platforms got easier to buy. The work to run them did not shrink.
What a loyalty programme actually costs in staff time#
Let us be specific. Below is a realistic breakdown of monthly hours for a single-brand hospitality operator running an in-house loyalty programme well. Not perfectly. Just well enough to drive incremental revenue.
| Activity | Monthly hours | Who usually owns it |
|---|---|---|
| Member communications (writing, scheduling, QA) | 8 to 12 | Marketing manager |
| Campaign planning and creative briefs | 6 to 10 | Marketing manager |
| Segmentation and list management | 4 to 6 | Marketing manager or data analyst |
| Reporting on repeat rate, AOV, incremental revenue | 4 to 8 | Marketing manager or founder |
| Staff training and refresh on enrolment | 2 to 4 | Operations |
| Vendor and tech management | 2 to 3 | Marketing manager |
| Total | 26 to 43 hours | Mostly one person |
At a fully-loaded marketing manager cost of around £45 an hour in the UK (salary plus on-costs), that is £1,170 to £1,935 a month of staff time. The software underneath might be £150 to £400. The platform is 10 to 25% of the real bill.
The four cost categories every operator should model before launch#
If you are about to launch a programme, or already have one and want to know what it really costs, model these four categories honestly. The build versus buy analysis from Propello covers the platform side. These four are what comes after.
- Platform and reward liability. Software subscription, plus the discount or freebie value redeemed by members. The easy one to model.
- Member communications. Hours to write, design, schedule and report on every push, email or wallet update that goes to members. Usually understated by half.
- Campaign and content development. Briefs, creative, copy, approvals. A monthly slow-Tuesday campaign is a real piece of work, not a five-minute task.
- Measurement and iteration. Pulling repeat rate, AOV by member versus non-member, cohort retention. Without this, you cannot prove the programme is paying for itself, which is how programmes get cut at the next budget review.
If you want a starting point, you can model the true cost of running loyalty in-house against the alternative of a managed programme. The gap is usually larger than operators expect, in both directions: in-house costs more than the invoice suggests, and a managed programme generates more repeat revenue because it is actually being run.
Why 'set it and forget it' kills repeat revenue instead of building it#
There is a quiet assumption in how loyalty software is sold: that once the programme exists, customers do the work. They enrol, they earn, they come back. The platform takes care of the rest.
This is not what happens. Benchmarking data across travel and hospitality loyalty shows that engagement collapses without active operating. Members enrol on a Friday night, get one stamp, and never hear from the brand again. Three months later they have forgotten the programme exists. The wallet pass sits in their phone, silent.
The metric that matters is not how many members you have. It is what share of them are active in a given window. We have written more on why enrolment numbers lie and active member rate is the only metric that matters, but the short version: a programme with 10,000 enrolled and 4% active is worth less than a programme with 2,000 enrolled and 35% active. The first is a database. The second is a revenue line.
The difference between owning a platform and running a programme#
This is the distinction the hospitality loyalty market has not made out loud. A platform is software. A programme is the platform plus the person running it, plus the campaigns, plus the segmentation, plus the measurement, plus the iteration. Most vendors sell the first and quietly assume you will handle the second.
| Owning a loyalty platform | Running a loyalty programme | |
|---|---|---|
| What you pay for | Software access | Outcomes: repeat rate, LTV, incremental revenue |
| Who briefs campaigns | Your marketing manager | A loyalty team that does this every day |
| Who writes member comms | Your marketing manager | Done for you |
| Who reports on impact | Your marketing manager (when they have time) | Reported monthly against revenue metrics |
| What happens if your marketer leaves | Programme stalls | Programme keeps running |
| Typical fully-loaded cost | Software fee + 25 to 40 staff hours per month | Single managed fee |
The piece on repeat revenue versus loyalty software goes deeper into this distinction. The short read: software is a tool. Tools do not produce outcomes on their own.
What changes when someone is actually running it#
When a programme has an operator, the same software produces a different result. The slow Tuesday gets a campaign. The lapsed three-month members get a wallet push. Tier benefits get tested and revised. The founder sees a monthly report that ties loyalty activity to repeat visits, not enrolment counts.
This is what Carrott does. We build the programme on wallet-native cards that live in Apple Wallet and Google Wallet, with no app for your customers to download. Then we run it: campaigns, member communications via the wallet pass, reporting against repeat rate and AOV. Your marketing manager gets their week back. You can read how Carrott runs the programme end to end, or look at how we grew repeat revenue for hospitality clients.
How many hours a month does an in-house loyalty programme actually take?
For a single-brand hospitality operator running the programme properly, expect 25 to 40 marketing hours a month covering member communications, campaign planning, segmentation, reporting and tech management. Multi-site groups tend toward the higher end.
Why do loyalty business cases usually underestimate the true cost?
Because they model the software fee and the reward liability, but not the staff time required to operate the programme. CFO-ready loyalty modelling shows operations and marketing costs are typically underestimated by 30 to 50%.
What is the difference between active members and enrolled members?
Enrolled members signed up once. Active members have earned, redeemed or engaged in a defined recent window, usually 30 to 90 days. Active member rate is the metric that correlates with repeat revenue. Enrolment alone does not.
Can we run a loyalty programme with no dedicated marketer?
Technically yes. Practically, the programme will under-perform, because nobody will brief campaigns, write win-backs or segment lapsed members. This is why managed or done-for-you models exist in the market.
How do we know if our existing loyalty programme is paying for itself?
Compare repeat visit rate and average order value of members against non-members over the same window, then deduct reward liability and fully-loaded operating cost (software plus staff hours). If the spread is positive and growing, the programme is working. If you cannot pull that report, the programme is not being measured, which usually means it is not being run.