Loyalty programs get sold on a long list of benefits. Some of them are real and measurable. Others are marketing copy that falls apart when you try to check it.
Here is the split.
The benefits that hold up
Existing customers come back more often
This is the core effect and the one worth building around. A customer sitting on seven of ten stamps has a reason to choose you over the place next door that they did not have last week.
The mechanism is not loyalty in any emotional sense. It is that you have created a small, concrete cost to defecting: the progress they would abandon. Behavioural researchers call this the endowed progress effect, and it is why a card that starts with two stamps already filled outperforms an identical card that starts empty.
The size of the effect depends almost entirely on how reachable the reward is. Set it eight weeks out and it changes behaviour. Set it eight months out and it does not.
You find out who your regulars are
Most independent owners can name their top five customers and are wrong about the rest. Not through carelessness. It is genuinely hard to track four hundred people across a busy counter.
A program turns that into a list. You can see who came four times last month and has not been back since, who only ever visits on a Saturday, and which customers are worth calling when you have something new.
That list is what makes everything downstream possible. Without it, "we should win back our lapsed customers" is a sentiment rather than a task.
You get a channel you own
Once someone joins, you can reach them. Not through a feed that shows your post to 4 percent of followers, and not by paying for it again each time.
This is the benefit that is easiest to underrate and hardest to replace. A cafe with six hundred members can fill a quiet Tuesday by sending one message. The same cafe with six hundred Instagram followers mostly cannot.
Use it carefully. The channel degrades fast if you send weekly discounts, and every message that is not worth opening costs you a little of the attention you will want later.
Staff get a script
A small one, but real. "Are you collecting stamps?" is an easier thing for a nervous eighteen year old to say than "would you like to hear about our offers". It gives the interaction a shape.
The benefits that are oversold
"Loyalty programs bring in new customers"
Mostly they do not. People do not choose a coffee shop because it has a stamp card. They choose it for location, quality, and habit, then join the card once they are already there.
A program can support acquisition if you attach a referral offer to it, but that is a different mechanic doing the work. Judge your program on frequency among existing customers and it will look effective. Judge it on new faces and it will look like a waste.
"It increases customer lifetime value by X percent"
You will see a lot of confident percentages. Treat them as marketing. The effect size varies enormously by trade, by reward design, and by whether staff mention it, and a figure from a US retail chain tells you nothing about a salon in Leeds.
What you can do is measure your own. Compare visits per member in the two months before and after joining. That number is yours and it is real.
"Customers feel valued"
Sometimes. A program is a discount with a delay attached, and customers know that. The feeling of being valued comes from staff recognising someone and knowing their order, which is a hiring and training outcome rather than a software one. A program can support it by showing staff who a regular is. It cannot manufacture it.
"It differentiates you from competitors"
It did in 2008. Now most of the shops on your street have something. A program is close to table stakes in hospitality, which means not having one is a small disadvantage and having one is not much of an advantage. The differentiation is in how well it is run, not whether it exists.
The costs to weigh against all this
Every completed card is margin you have given away, and some of it goes to people who would have come anyway. That is unavoidable and it is the price of the mechanic. What you control is the threshold.
There is also an attention cost. A program is one more thing for staff to remember during a rush, one more thing to explain, one more thing that can go wrong at the till. If the program is fiddly, staff will quietly stop offering it, and a program nobody offers is pure cost.
We broke the numbers down in how much a loyalty program costs.
How to tell whether yours is working
Four numbers, checked monthly.
Signup rate: what share of transactions involve a member. If it stays under about a quarter after two months, staff are not mentioning it.
Visit frequency before and after joining, per member. This is the number that tells you whether the program is doing its job.
Redemption rate: the share of completed cards that get claimed. Low redemption is not a saving. It means people stopped caring before they finished.
Reactivation: how many lapsed members come back after you contact them. This is the return on having a channel.
If the first two look healthy and the last two do not, your reward is fine and your follow up is missing. Understanding customer health segments covers what to do about that.
Where to start
If you are still deciding whether to run one at all, what a loyalty program is covers the mechanics without the sales pitch. If you have decided and want the build order, go to how to create a loyalty program.