Customer Loyalty Program: The Operator's Guide to Building One That Pays

Most of the time, when an operator tells me they are launching a customer loyalty program, what they are actually launching is a discount program with a nicer name. The points are a currency. The currency buys money off. And the people who earn the most of that currency are the clients who were already coming in every four weeks, at full price, without any of it.

I am not against loyalty programs. We have clients running good ones. But I have looked at enough multi-location P&Ls to know that a loyalty program is one of the easiest things in this industry to launch badly and one of the hardest to measure honestly. So this guide is written the way I would talk it through on a call: what these programs are, where they earn their keep, the arithmetic that decides whether yours makes money, and the specific situation where a loyalty program is the wrong tool entirely.

If you want the broader context first, our customer retention strategies for service businesses page is the pillar this sits under. A loyalty program is one lever on that list, not the list.

What a customer loyalty program actually is

Strip the branding off and a loyalty program is a standing offer: the business gives up a slice of margin on future purchases in exchange for a change in customer behaviour, plus the data that comes with identifying who is buying.

That is the whole mechanism. Three parts:

A rule. Spend this, visit this often, stay a member this long, and you earn something.

A reward. Money off, a free service, early access, a tier badge, a product credit.

An identity. Every transaction gets attached to a person, which is worth something on its own if your point-of-sale or booking system was previously recording half your visits as anonymous walk-ins.

The reason operators underestimate the cost is that the rule and the reward look free at launch. Nobody has earned anything yet. The liability builds quietly over the first two quarters, and by the time redemptions start landing, the program has usually been declared a success on enrolment numbers alone.

The honest case for and against

Here is my actual opinion, and it has two halves.

The case for

Identification is genuinely valuable. The single most expensive gap in your data is the client you cannot name. If a loyalty program takes your identified-transaction rate from 60% to 95%, you now have a CRM that can support segmentation, recall workflows, and lapse detection. That alone can be worth more than the reward cost, and it is the benefit operators talk about least.

Frequency programs fit certain service lines well. Where the service has a natural repeat interval and clients genuinely drift between providers, a small nudge back to you rather than the studio near the new office can pay for itself. Blow-dry bars, nail salons, quick-service wellness, injectables that need maintenance.

Tiers can lift average ticket. When the next tier is visible and close, some clients add a service to reach it. That is real incremental revenue, not just a discount on planned spend.

The case against

Most programs subsidise behaviour you already had. This is the big one. Your best clients enrol first and redeem most. If a client was coming in every five weeks before the program and every five weeks after, every point she earned is pure margin given away. Loyalty spend concentrates on the least at-risk part of your base, almost by design.

It becomes a permanent price cut you cannot reverse. A promotion ends. A loyalty program does not. Once clients have earned a balance, withdrawing it is a customer-service event across every location you run. You are committing to a lower effective price indefinitely, and you should price it that way from day one.

It competes for the same attention as the things that actually retain. In service businesses, rebooking at checkout, provider continuity, and following up on a missed appointment beat any points balance. I have watched teams spend a quarter of operational bandwidth on a loyalty launch while the front desk was still not asking for the next booking before the client left. That trade is almost always negative.

Redemption is not the same as loyalty. A high redemption rate proves people like free things. It does not prove anyone visited more often than they would have. The only number that matters is incremental lift, and most programs are never measured that way. More on that later.

So: worth doing when identification is poor, when your category has real switching, and when you have the discipline to model the margin before launch. Not worth doing as a response to rising churn. Churn has causes, and a points balance is not a cause. Our guide to what customer churn is covers how to find the real ones.

Program types and the trade-offs that matter

There is no best structure. There is a structure that fits your repeat interval, your margin, and how much your front desk can realistically explain in fifteen seconds.

Points

The client earns points per dollar and spends them on rewards. Flexible, familiar, and the easiest to get wrong, because the point is a currency you are inventing and you have to decide what it is worth.

Points suit businesses with variable ticket sizes: med spas, salons with a wide service menu, retail-heavy locations. They suit high-frequency, low-variance businesses badly, because they add accounting to something a punch card handles for free.

The trade-off: flexibility for complexity. Points create a balance sheet liability, need software, and require staff to answer “how many points do I have?” all day.

Tiers

Clients move up levels based on spend or visits over a rolling window, with better benefits at each level. Tiers work when status has social value and when your menu has upsell headroom. They are strong in med spas and premium fitness.

The trade-off: tiers only motivate the people near a threshold. The client sitting comfortably in the middle of a tier feels nothing. And demotion is a real problem. Dropping a long-term client a level for missing a spend target is a great way to trigger the churn you built the program to prevent.

Paid or VIP membership

The client pays a recurring fee for ongoing benefits: a set number of services, member pricing, priority booking. Strictly speaking this is a recurring-revenue product rather than a loyalty program, and it is the structure I like most for service businesses.

Why: it takes money instead of giving it away, it creates predictable revenue, and paying for membership is a commitment that changes behaviour in a way earning points does not. It also gives you a real cancellation signal, which a points program never will.

The trade-off: you have to be worth the fee every month, and you now have involuntary churn to manage when cards fail. That is its own discipline, and we have written about dunning and failed payments separately.

Punch or visit-based

Ten visits, the eleventh is free. Unbeatable for simplicity. Every client understands it instantly and every staff member can explain it.

Best fit: consistent-price, high-frequency services. Blow-dry, waxing, nails, spray tan, class packs.

The trade-off: it rewards volume, not value, and the free visit lands at full cost to you. Run the margin on the eleventh visit before you print the cards. If your service line is 60% gross margin, a one-in-eleven giveaway is roughly a 9% price cut on everyone who completes a card.

Cashback or account credit

The client earns a credit balance redeemable against future services. Simpler than points because the value is obvious, and credit locks spend inside your business rather than handing over cash.

The trade-off: it is a discount with no ambiguity, so you get less perceived-value lift per dollar given up than a program where the reward feels like a gift.

Referral hybrids

Reward for bringing someone in, usually credit for both sides. Structurally this is acquisition, not loyalty, but it often ships inside the same program.

Referral is the one hybrid I actively encourage, because you are paying for a new client rather than discounting an existing one, and you can compare the cost directly against your paid acquisition cost. Just cap it and require the referred client to complete a paid visit before either reward unlocks.

Matching structure to model

Roughly how I think about fit:

Business model	Usually fits	Usually does not
High-frequency, flat price (waxing, nails, blow-dry)	Punch or visit-based	Points
Wide menu, variable ticket (med spa, full-service salon)	Points, or tiers	Punch cards
Class or session based (boutique fitness)	Paid membership, prepaid series	Cashback
Appointment-driven clinical (dental, DSO)	Recall discipline, not loyalty	Points
Ecommerce and subscription	Points or tiers, tied to reorder interval	Punch cards

Dental is the one I will state flatly. In a dental group, a points program is almost never the answer. The revenue is in hygiene recall and treatment-plan follow-up, and those are operational problems.

Loyalty program examples

Useful examples come in two flavours: the large public programs everyone has seen, and the quieter patterns that work at multi-location service scale. The second group is where I would spend your attention, but the first is worth understanding because it is what your clients compare you to.

Well-known programs, at common-knowledge level

Starbucks Rewards is the reference point for points done at scale. Order through the app, earn stars, redeem for drinks and food. The real lesson is not the points. It is that the program is fused to the ordering experience, so enrolment is not a separate ask. The program and the transaction are the same action.

Sephora’s Beauty Insider is the standard tiered example in beauty. Spend-based tiers, product rewards, and early access to sales. The lesson is that the non-discount benefits, access and exclusivity, carry a lot of the perceived value. Worth noting for any operator who assumes rewards have to be money off.

Amazon Prime is the paid-membership archetype. You pay up front and the benefits make you buy more inside the ecosystem. That is the structure most service operators should be studying, not points.

Airline and hotel programs are the origin of tiering and also the cautionary tale: decades of devaluation and earn rules members track more closely than the operator does. If you build tiers, build them so you never need to devalue them later.

Do not copy any of these mechanics without running your own numbers. The margin structure of a coffee chain has nothing in common with a med spa.

Service-business patterns that work

These are the ones I see actually move revenue in multi-location operations.

The prepaid series with a built-in bonus. Buy six sessions, get a seventh. You take cash up front, you lock in the visit commitment, and the bonus session costs you variable cost rather than full price. This is a loyalty rewards program dressed as a package, and it is the most operator-friendly version I know.

The membership that pauses instead of cancels. For gyms and studios, the single highest-value benefit you can offer is a graceful pause. Clients who pause come back at meaningfully better rates than clients who cancel, in our experience across fitness clients. Making pause a member benefit turns a retention mechanic into a reward.

Tier benefits that cost you nothing marginal. Priority booking windows, first access to a new injector’s calendar, a dedicated line for rescheduling. These have real value to clients and near-zero margin cost to you. The most efficient loyalty benefit is one your operation can produce for free.

Provider continuity as a reward. In salons and med spas, the relationship is with the person. A benefit that guarantees the same provider, or gets a client onto a preferred provider’s book faster, attacks the actual cause of churn.

Enrolment tied to rebooking. Make the program’s first reward unlock only when the next appointment is on the calendar. Now your loyalty program is doing the one thing that predicts retention best. See the salon and spa and fitness pages for how this plays out by vertical.

Designing the economics, which is where most operators go wrong

This is the section to read twice. Everything below is a model you fill in with your own numbers. I am not going to quote industry benchmark redemption rates as fact, because the published ranges vary enormously by category and most of what circulates online is unsourced.

Decide what a point is worth before you decide how points are earned

Work backwards. Pick the discount rate you are willing to run permanently. If that is 5% of revenue, then the entire reward structure has to fit inside 5% of revenue at full redemption, not at the redemption rate you hope for.

The common mistake is setting earn rates first because they sound generous, then discovering the implied discount is 12%.

Model it on margin, not revenue

A worked example, with the assumptions labelled so you can swap yours in:

Illustrative model, not a benchmark. Assume 2,000 active clients, average ticket $120, average 6 visits a year, so $1.44M annual revenue. Assume 65% gross margin on service. The program gives 5% back in credit, earned on all spend.

Full redemption cost: $72,000 a year, all of it out of a $936,000 gross margin pool, so roughly 7.7% of gross margin.

Now assume 70% of credit is redeemed and 30% expires. Cost drops to about $50,400.

For the program to break even, it has to generate about $77,500 in additional revenue at 65% margin. That is roughly 646 extra visits a year, or about 0.32 extra visits per active client per year.

Is 0.32 extra visits per client achievable? Maybe. That is the question the model is for. What matters is that you asked it before launch rather than eighteen months in.

Breakage is real but do not build on it

Breakage is the portion of earned rewards that never gets redeemed. It makes programs look cheaper than they are, and it is the reason expiry policies exist.

Two cautions. First, breakage that comes from clients forgetting is fine. Breakage that comes from clients giving up because the reward was unreachable is churn with extra steps. Second, expiry rules on stored value are regulated in a lot of places and the rules vary by state and country. Talk to your accountant and your lawyer before you set an expiry policy, especially if you operate across state lines.

Put the liability on the balance sheet

Unredeemed rewards are a deferred obligation. If you are a multi-location group with any kind of outside investor, lender, or eventual sale process, someone will eventually ask what the outstanding loyalty balance is. Track it monthly from day one. Retrofitting that number from transaction logs two years later is miserable work.

The metrics worth tracking

Enrolment rate by location, because it tells you which front desks are actually doing it.
Active participation rate, meaning enrolled clients who earned or redeemed in the last 90 days. Enrolment alone is vanity.
Redemption rate and outstanding balance, for liability.
Incremental visit frequency for enrolled versus comparable non-enrolled clients. This is the one that says whether the program works.
Margin per redemption, so you know the real cost of each reward event.

For retention and churn measurement generally, see our guides on customer retention rate and how to reduce customer churn.

Loyalty program software and management

The loyalty program software category does a fairly narrow set of jobs: it tracks earn and burn, holds the client balance, handles enrolment and identification, sends the related comms, and reports on liability and redemption. Some tools add tiering logic, referral tracking, and integrations into point-of-sale.

Before you buy anything standalone, check what you already own. The booking and CRM platform most multi-location service operators already run, the one that holds the schedule and the client record, very often includes loyalty, points, memberships, or packages in a module you are already paying for. Turning that on is nearly always better than adding a separate system, because a loyalty balance that lives outside the booking platform is a balance your front desk cannot see at checkout.

Evaluation criteria I would use, in priority order:

Native to your booking or point-of-sale system, or with a real two-way integration. Not a nightly CSV.
Visible at checkout. If staff cannot see the balance on the screen they are already using, the program will not be mentioned.
Liability reporting. If it cannot tell you the outstanding balance by location, it is not finished software.
Multi-location logic. Can a client earn at one location and redeem at another, and does the reporting attribute the cost to the right P&L? Franchise groups, this is the question that causes fights.
Cohort reporting, so you can compare enrolled and non-enrolled clients rather than only counting redemptions.
Client-side simplicity. If explaining the rule takes more than one sentence, adoption dies at the front desk.

I am deliberately not reviewing individual tools here. We covered the broader category in our customer retention software guide, which is the right place to start if you are still building a shortlist.

On loyalty program management as an ongoing job: give it an owner. Not a committee, not the marketing calendar. One person who reviews enrolment by location, outstanding liability, and incremental lift every month, and who has the authority to change earn rates. Programs without an owner drift, and drifting programs always drift toward giving more away.

Launching and running it

Enrol at the point of sale, verbally. Not by QR code on a poster. The single biggest determinant of enrolment rate is whether the person at checkout asks. Script it, keep it to one sentence, and audit it.

Incentivise the staff, carefully. Pay for enrolments and you will get junk enrolments. Pay for enrolled clients who complete a second visit and you get the behaviour you want. Tie any staff incentive to the downstream outcome, not the signup.

Keep comms light. A welcome message, a balance update when something meaningful changes, and a nudge when a reward is close or about to expire. That is it. A loyalty program that generates weekly email is a loyalty program clients mute.

Measure incremental lift, not gross redemptions. Hold out a comparable group if you can, or at minimum compare visit frequency for enrolled clients in the 12 months before and after enrolment against a matched non-enrolled cohort. Gross redemption value is the number that makes a loyalty program look successful and tells you nothing.

Review earn rates annually. Not to devalue, but to confirm the margin model still holds after price changes and service-mix shifts.

When a loyalty program is the wrong answer

Here is the situation I see most, and it is the reason this section exists.

An operator is losing clients. Monthly actives are drifting down across the group. Someone proposes a loyalty program, and a quarter of operational capacity goes into designing it.

The problem is that a loyalty program only reaches people who are still walking through the door. It has no mechanism for the client who last visited eleven months ago. She is not going to see a poster, she is not opening the emails, and she has no balance to care about. Meanwhile the database of lapsed clients keeps growing, and for a multi-location group it is routinely the largest single pool of recoverable revenue on the P&L.

Loyalty is a retention motion for active clients. Reactivation is a separate motion for lapsed ones, and the two need different systems. Our customer reactivation guide covers the second one properly, and the ROI calculator will tell you what your own inactive list is worth against your average ticket.

The sequencing I would recommend: fix rebooking at checkout first, build the reactivation motion second, launch the loyalty program third. Most groups do it in reverse, and then wonder why the program did not fix churn.

For what it is worth, that is also where our business sits. Winback Engine runs the third motion for multi-location operators. Your automated flows run first, and trained human agents call the clients those flows did not recover. We price on the revenue actually recovered, which keeps everybody honest about incremental lift, a discipline I would apply to a loyalty program too. Worth noting for med spa operators in particular, where ticket values make the lapsed list unusually valuable.

FAQ

What is a customer loyalty program?

A customer loyalty program is a structured offer that rewards existing clients for repeat purchases, visits, or spend, usually with points, tiers, credits, or free services. For operators, it trades a defined amount of future margin for higher visit frequency and, just as importantly, for identifying more of your transactions against named client records in your CRM.

Do customer loyalty programs actually increase retention?

Sometimes, and less often than operators expect. The risk is that a program rewards clients who would have returned anyway, which costs margin without changing behaviour. The way to know is to compare visit frequency for enrolled clients against a matched group of non-enrolled clients, rather than counting redemptions. Programs that tie the first reward to booking the next appointment tend to perform better than open-ended points.

What type of loyalty program is best for a service business?

For high-frequency flat-price services such as waxing or blow-dry, a visit-based punch structure is simplest and adopts fastest. For wide-menu businesses such as med spas and full-service salons, points or tiers fit better. For studios and gyms, a paid membership or prepaid series is usually stronger than any earn-and-burn program, because it takes revenue up front rather than discounting it later.

How much should a customer loyalty program cost?

Decide the permanent discount rate you can sustain, model it at full redemption against gross margin rather than revenue, and build the earn rates to fit inside that number. Do not rely on unredeemed balances to make the economics work. Then calculate how many additional visits per client per year the program must generate to break even, and judge whether that figure is realistic for your category.

What should I look for in loyalty program software?

Prioritise software that is native to the booking or point-of-sale system your locations already run, so balances are visible at checkout. Beyond that, require liability reporting by location, multi-location earn-and-redeem logic with correct P&L attribution, and cohort reporting that compares enrolled and non-enrolled clients. Check what your existing platform already includes before buying a standalone system.

Should I run a loyalty program or a reactivation campaign first?

Reactivation first, in most cases. A loyalty program only reaches clients who are still visiting. If a meaningful share of your database has not booked in six months or more, that list is usually the larger and faster source of recoverable revenue, and recovering it does not require permanently discounting your active clients.