Most customers do not leave angry. They drift. Nobody called, no reason to come back arrived, and on the day the need came round again somebody else happened to be in front of them. That is what churn looks like in a business with nothing to cancel: not a complaint, just a gap that gets longer until it is permanent.
The software is the easy part. Your point of sale probably sells a loyalty module already, and it does what it says: customers earn points, points become a discount. What it cannot tell you is who belongs in the program, what should trigger it, what the reward costs you in margin, or what counts as lapsed for your business. Those four decisions are the product. We design them, then wire them into the systems you already run.
How do you know you have a retention problem?
The hard part of churn is that nothing happens. A cancelled subscription sends an email; a customer who quietly stops coming sends nothing at all. So the first question is never how many you lost, it is what normal looks like here. A restaurant regular might come every week, a salon client every six weeks, a wholesale account every month like clockwork. Until you know that rhythm, there is no difference between a customer who is busy and a customer who is gone.
Do loyalty programs actually work?
Somewhere in the pitch you will be shown a number: raise retention 5% and profits rise 25 to 95 percent. It is the most quoted statistic in this industry, and it is worth knowing where it came from. The original study, Reichheld and Sasser in the Harvard Business Review, September 1990, reported that cutting defections by 5% generated 85% more profit in one bank's branch system, 50% more in an insurance brokerage, and 30% more in an auto-service company. The 25% floor comes from a different document, a Bain brief from October 2001, which says plainly that it describes financial services. The 95% is in neither.
Read that again with your own business in mind. Of the three companies in the original research, the one that resembles a local service business, the auto-service chain, produced the smallest number in the set. The figure being quoted at a Pensacola contractor in 2026 was measured inside a bank in 1990. A number repeated often enough starts to feel like evidence, and this one is not. Retention is worth doing, and it does not need a borrowed statistic to justify it.
Here is what the evidence does support. A meta-analysis of forty years of research published in the Journal of the Academy of Marketing Science in 2022 pooled 429 measured effects from studies between 1990 and 2020. Loyalty programs do work, with two qualifications that matter more than the headline: they change what customers do far more readily than how customers feel, and how well they work depends systematically on how they are designed. A program is not a thing you install, then. It is a set of decisions about who, when, and what for.
THE LOCAL ANGLE
Seasonality makes this harder here than the textbooks assume. A business with a real summer, or one that empties out the week after spring break, has a normal gap that moves through the year, so a win-back campaign running on a fixed calendar chases people who were never gone and misses the ones who were. Storm weeks break everybody's rhythm at once. We set the window against your own history instead of an industry benchmark.
What's included in a retention program?
How the design actually works
Set the baseline
What normal looks like for each kind of customer, what counts as lapsed, and what a repeat customer is actually worth to you. Without those three numbers everything after it is guesswork.
Design the play
Who the program is for, what triggers it, what the reward is, and what it costs in margin. The design works backwards from the reason people leave, which is why finding that out comes first.
Wire it and watch it
The triggers get built into the systems you already run, then measured against each customer's own prior rhythm. Timing runs on the system; the words stay yours.
What loyalty is not, and what has to come first
Loyalty and churn work sits in Retain and it compounds, worth more in year three than in year one, because it works on a relationship that already exists. It leans on three things. Customer journey and friction audits find out why people actually leave, which is the input this service is designed around. The CRM holds the history that turns lapsed into something computable rather than a hunch. Email marketing is how the messages go out, sending from the CRM so the timing follows the customer instead of the calendar.
What this is not is a punch card. A punch card is a mechanic, and a mechanic is not a reason. Ten stamps for a free coffee works when the coffee was already good and the shop was already convenient; it does nothing for a business people are leaving for a reason nobody has identified. A blanket discount dressed up as a program fails in reverse: it hands margin to the customers who were going to buy anyway and teaches everyone else to wait for the offer. Retention is an experience problem before it is a program problem.
So this is the call we will make even when it delays the work: launching a loyalty program before you know why people leave subsidizes the customers who were already staying. It looks like progress, because sign-ups climb. The customers who were drifting are still drifting, and now they are drifting with a discount.
What does a loyalty program cost?
This one is scoped per engagement rather than sold at a starting price, and the reason is a single variable: how many distinct kinds of customer you have. A business with one repeat cycle is a different build from one with four, and the design scales with that rather than with a package tier. So the number comes out of a short conversation about your customer base, not off a list. The software that runs the program is separate, billed to you and in your name, and most businesses already pay for something that can do the job.
- Who it isn't for
- If your business has no repeat cycle, this is not for you yet, and some genuinely do not: a roofer whose customers need them twice in twenty years is running an acquisition business, not a retention one. It is also wrong when almost nobody is arriving in the first place, because retention multiplies what you already have, and multiplying a very small number gives you a very small number. And if people are leaving over the product or the service itself, a program will not outrun that.
- What you own
- You own the program: the segments, the rules, the templates, and the history it builds. The software subscription is in your name rather than ours, so if we part ways it keeps running and none of it lives only with us.
Still reading? Then the number didn't scare you off, and the free live look is worth thirty minutes: your setup, looked at live, and the notes are yours to keep.
Book a Friendly CallGOOD QUESTIONS
Is a loyalty program just a punch card?
A punch card is one mechanic, and for a coffee shop it can be the right one. What it cannot do is tell you why somebody stopped coming, and for most businesses that is the question worth money. For a service business the answer usually looks like a maintenance plan, priority scheduling in the busy season, or first access to something worth having. We design the mechanic to the business model; the point is a standing reason to come back.
Do loyalty programs actually work?
Yes, with two qualifications worth knowing before you spend anything. The largest review of the evidence, a meta-analysis of 429 measured effects from studies published between 1990 and 2020 in the Journal of the Academy of Marketing Science, found that programs do increase loyalty, that they shift what customers do much more readily than how customers feel, and that how well they work depends systematically on how they are designed. A program works when it is designed for the business it runs in. One copied off a competitor's counter usually does not.
How do you measure churn when nothing gets cancelled?
You define the gap first. For every kind of customer there is a normal interval between visits or orders, and lapsed is that interval plus a margin you set deliberately. Once that exists, churn stops being a feeling and becomes a number you can watch weekly, and you can act on it while the customer is still winnable. Most churn advice online is written for subscription businesses, where a cancellation announces itself. A shop or a trade never gets that signal, so the window has to be built.
Where does a referral program fit into this?
A referral program is a retention play aimed at customers you already have, so the mechanics belong here: who gets asked, when, what the incentive is, and how it gets tracked. What the referral produces afterwards, the review, the testimonial, the photo of somebody using the thing, is social proof work and lives on that page. The handoff matters in both directions, because a referral asked for at the wrong moment produces nothing, and one nobody captures produces nothing you can show.
Ready to go deeper on Loyalty & Churn Programs?
Every engagement starts with a free live look, thirty minutes, honest notes, no strings.
Book the Free Live Look