In brief
A ranked list of vendors would be out of date by spring. This is the method instead: what the four kinds of free actually cost, how to budget effort rather than money, and what a two-week trial can genuinely prove.
The short answer
With no IT person, the deciding number is not the subscription. It is the total effort: the hours to set it up, the minutes to train each member of staff, and the minutes a week it will take forever after. A free tool that costs an hour a week costs about £52 a month of somebody’s time at £12 an hour, which is more than five times a £9 subscription that costs none.
So evaluate in this order. Work out what “free” means in each case, add up the effort, then run a two-week trial designed to test the counter rather than the mechanic. Settle the export and exit questions before you start, not when you want to leave.
Why this is a method and not a list
A ranked list of the ten best loyalty tools for small business would be more comfortable to read and would be wrong within a couple of quarters. Prices change, free tiers move, and the shop the list was written for is not yours.
A method survives that. It also produces an answer you can defend to yourself in six months, which matters more than it sounds when the program is annoying somebody at 8am and you have to remember why you picked this one.
The other reason: the thing that fails is almost never the software. It is the twenty minutes a week nobody has. That is not visible on a comparison page, and it is the single best predictor of whether you will still be running the program at Christmas.
The four things “free” means
The word does very different work depending on who is using it, and only two of the four survive contact with a real shop.
| What is offered | What you actually have | Usable long term |
|---|---|---|
| A free trial | The full product for a fixed window, then a card | No. It expires on a date, not on a decision |
| A free demo tier | Everything configurable, the customer-facing part switched off | No. It is a sales tool, not a plan |
| A free tier with a cap | A working program, bounded by a number | Yes, if the cap is the right number |
| Free forever, uncapped | A working program with no visible meter | Yes, but find out who is paying for it |
The two usable ones are the capped tier and the genuinely free product, and each has a follow-up question.
For a capped tier, ask what is capped and what happens at the cap. A cap on customers is fair, because it rises with your success and the vendor’s costs rise with it. A cap on stamps or scans is a toll on the operation you were trying to speed up, and it bites hardest on your best day. Ask whether hitting the cap stops new joins or freezes everyone, because the second punishes customers for your billing.
For free with no cap, ask what pays for it. Advertising, data resale, or a marketplace that puts your competitor next to you are all real business models, and none of them are disqualifying if you know about them. The full breakdown, including what happens to your list at each kind of limit, is in what free actually buys you.
The total-effort budget
Three numbers, and the third one is the one that decides it.
Setup hours, once. Creating the card, writing the rules, printing the poster, adding staff. Time this in a trial rather than estimating it. Anything above a working day for a single-site stamp card is a signal that the tool was built for a different size of business.
Training minutes, per person, repeatedly. Not once. Staff change, and the tool has to be learnable by whoever starts next month without you being there. The realistic test is whether a new starter can issue a stamp correctly after being shown once.
Minutes a week, forever. Reconciling counts, fixing a wrong stamp, matching a phone number to a customer, explaining to someone why their card says four when they say five. This is where a cheap tool gets expensive.
Here is the arithmetic, with the assumptions on the page. Value your own time at £12 an hour, the same as a counter wage, which understates it if you are the owner.
| Free tool, one hour a week | £9 tool, five minutes a week | |
|---|---|---|
| Subscription, per year | £0 | £108 |
| Weekly minutes, per year | 52 hours | 4.3 hours |
| That time, at £12/hour | £624 | £52 |
| Year one total | £624 | £160 |
The free tool costs nearly four times as much, and it charges in the currency you have least of. Change the wage or the minutes and the direction of the answer does not move, because an hour a week is 52 hours a year and no subscription in this category costs that.
This is separate from what the program itself costs, which is dominated by the rewards you hand over rather than by any of the above. That model is in what a loyalty program actually costs.
A two-week trial that tests something
Start by accepting what two weeks cannot tell you. A ten-stamp card at one visit a week takes ten weeks to complete, so a fortnight ends before your first customer has finished one. The trial cannot test whether a loyalty program works for your shop. It can test whether this software survives your counter, which is the part that actually fails.
Seven things to do inside the window, in roughly this order:
- Join it yourself, on your own phone, standing where the customer stands. Count the taps between the poster and the first stamp. Do it once more with your reading glasses off.
- Hand it to your newest member of staff with no training and watch them issue a stamp. Say nothing. Whatever they get stuck on is what everyone will get stuck on.
- Redeem a reward during your busiest twenty minutes. Redemption is where programs die, and a quiet Tuesday will not show you that.
- Try to cheat it. Scan the poster twice. Ask for two stamps in one visit. See whether the product stops you or whether the person behind the counter has to.
- Turn the shop wifi off and take a stamp on mobile data, or on nothing. Basements and bad Saturdays are normal.
- Change the reward on day ten and look at a card that already has stamps on it. This is the single most revealing test in the list, because you will change the reward eventually.
- Export everything on day thirteen and open the file. Not “request an export”. Download it, and check that customers, balances and redemption history are all in there.
Recruit five named regulars rather than testing on staff alone. Five real customers will find the thing you cannot see because you already know how it works.
The rest of the vendor questions, including the ones you cannot test and have to get in writing, are in ten things to check before buying.
Settle migration and exit first
Four questions, all easier to answer before you have anything to lose.
What happens to the paper cards already out there? Decide the policy before day one and write it on a card by the till. Honoring both for a fixed period is the usual answer, and it is a decision, not an oversight.
Do you have consent to import your existing list? An email address collected for order confirmations was not necessarily given for loyalty marketing. If in doubt, start the list fresh from the poster and let people join deliberately.
What is the export, exactly? A file format, containing balances, that you can download yourself. Get the answer before you sign, because the value of the answer is highest on the day the relationship is worst.
Who is the data controller? If it is the vendor rather than you, your customers are also theirs. That is a legitimate model for some products, and you should know which one you bought.
The worked example, with real numbers
A total-effort model needs a real price in it, so here is one, in full and once.
NeoLoyal is free for up to 30 customers, then £9 a month on Starter, £18 on Growth and £36 on Chain, billed in GBP. New businesses open on 30 days of Growth, and that period ends by dropping to the free tier rather than at a paywall, so nothing is switched off on a date. Current tiers and what each includes are on the pricing page, which will stay accurate longer than this sentence will.
Substitute any vendor’s numbers into the effort table above. The model is the point, and it works the same way with a figure that is not ours.
Two numbers, written down first
Before you open a single pricing page, write down how many minutes a week you will genuinely give this, and how many customers you expect to have on the card in a year.
The first number disqualifies most of the market immediately, because a tool that needs weekly attention is not a candidate if your honest answer is ten minutes. The second turns every pricing page into a single figure you can compare, instead of a tier chart you have to interpret. Then pick the mechanic before the vendor: planning a visit-based program is a shorter decision than choosing software, and it is the one that determines which software could possibly fit. If you want the category-level version of that question first, it is in loyalty programs for small businesses.