Most loyalty schemes in hospitality do one of two unhelpful things. Either they do nothing at all, a dusty stamp card that no one remembers to stamp, or they quietly train your best guests to wait for money off. The second is worse than the first, because it takes people who were happily paying full price and teaches them to expect a discount. You can lose margin and call it loyalty.
A loyalty programme done well is one of the most powerful tools an independent has, because it works on the part of your business with the most money in it: getting the people who already like you to come back more often. But the design is everything, and most of what is sold as loyalty software gets the design wrong. This is how I would think about it, having run schemes that did not work before building one that does.
What loyalty is actually for
Start with the goal, because most schemes are built without a clear one. Loyalty is not a discount mechanism. Its job is to change behaviour in your favour and to make your best guests feel recognised. Those are two different jobs and a good scheme does both.
Changing behaviour means nudging people to come more often, to come on the nights you need filling, and to spend a little more when they do. Making people feel recognised means that your regulars experience something your competitors cannot easily copy: the sense that this place knows them and values them. A discount does neither of those well. Anyone can discount. Recognition is the thing money cannot quickly buy.
If your scheme is not clearly doing one of those jobs, it is just giving margin away.
The discount trap
Here is the failure mode to avoid above all. A blanket discount, ten per cent off for members, say, feels generous and is mostly self-defeating. It rewards everyone equally, including the people who would have paid full price and the people who only ever come when there is an offer. It anchors your regulars to a lower price, so full price starts to feel like a penalty. And it attracts the least loyal customers of all, the ones who follow discounts from venue to venue and vanish the moment a better one appears.
The guests worth keeping are not chasing ten per cent. They come because they like your place, and what they want is to feel that liking returned. A small, well-timed, unexpected gesture, a drink on the house for a regular, a held table, a bit of recognition, buys more loyalty than a standing discount and costs you less.
Points and tiers: two tools, two jobs
The two building blocks of most schemes are points and tiers, and they are good at different things.
Points reward spend and let you nudge. Every pound earns something, and you can weight that, more points on a quiet night, bonus points for trying the new menu, a reward that unlocks at a level that encourages one more visit than someone would otherwise make. Points are your behaviour lever.
Tiers recognise value. Rather than treating every member the same, tiers let you quietly sort guests by how much they actually matter to your business and treat your champions differently from someone who joined once. The recognition does not have to be loud. Often the best version is invisible to the guest and visible only to your team: this person is one of our best, look after them.
The strongest schemes use both. Points to encourage the behaviour you want, tiers to make sure your most valuable people feel it. A scheme that only does points treats your best guest the same as your newest. A scheme that only does tiers has no lever to pull.
Why loyalty has to live on the bill
This is the part most loyalty products get wrong, and it is the reason so many schemes feel disconnected from the business.
A standalone loyalty app sees what it is shown. Usually that is a check-in or a scan, which tells it someone was here but not what they did. So it rewards visits, or it rewards a guess. The trouble is that a visit is a poor measure of value. The guest who pops in for a coffee and the guest who brings six people for dinner are not equally valuable, and a scheme that cannot tell them apart will reward the wrong behaviour.
Loyalty that lives on the till sees the whole bill. It knows what was actually spent, so it can reward real value automatically, without anyone scanning anything. The points are earned on the spend, the tier reflects genuine worth, and the guest does not have to remember to do anything. This is also why loyalty belongs in the same system as your guest record rather than in an app off to the side. I have written separately about why loyalty belongs in your till, not an app, because it is the single biggest design decision in the whole area.
The underused job: filling your quiet nights
Here is a use of loyalty that almost no one talks about and that pays for itself. Most venues have a demand problem that is really a shape problem: the weekend is rammed and the early week is dead. Discounting the quiet nights is the obvious answer and the wrong one, because it just trains people to wait for the cheap night and erodes your midweek price.
Loyalty does it better. Award bonus points, not a discount, for booking or visiting off-peak. The guest gets more reward for coming on a Tuesday, you give away something cheaper to provide than a price cut, and you have shifted demand to where you actually have capacity. You are using loyalty as a lever on the shape of your week, not as a blunt giveaway.
Consent, data and doing it properly
A loyalty scheme generates exactly the kind of guest data that makes the rest of your marketing work, which means it has to be run properly. Membership should sit on the single guest record, not in a separate silo, so that what someone earns and how they behave informs everything else you know about them. And the consent that comes with signing up should be tracked correctly, per channel, so that loyalty feeds your marketing without crossing a line. A scheme that hoovers up data and handles it loosely is a problem dressed as an asset.
How to choose a loyalty product
Take these questions into any conversation. Does it reward real spend across the whole bill, or only visits and check-ins? Does it tie to one guest record, or sit as a separate app? Can I use both points and tiers, or only one? Can I weight rewards to shift behaviour, like filling quiet nights? Does it give me control over who gets what, or only a blanket offer? And does it handle consent properly so it strengthens my marketing rather than risking it?
If a product can only offer everyone the same discount and only knows that someone showed up, it will cost you margin and call it loyalty.
Where Grace fits
In Grace, loyalty is a module of the platform rather than an app bolted on, so it sees the whole bill through the EPOS and ties to the same guest record as everything else. Every guest is classified by behaviour automatically, points can be earned on bookings and spend, rewards can be redeemed at the till, and you can award bonus points for off-peak bookings to fill the nights that need filling. Because it is one platform, what a guest earns and how they behave feeds the rest of what you know about them rather than living in a corner on its own. You can see the Loyalty module in detail, and the wider thinking on our restaurant loyalty software overview.
It works this way because I ran schemes that did not, and the difference was always the same: the ones that failed rewarded the wrong thing because they could not see the bill. If a feature cannot survive a Friday night, it does not ship, and a loyalty scheme that makes your team scan cards during service does not survive one.
The short version
Decide what your scheme is for, changing behaviour and recognising value, and build for that, not for discounting. Avoid the blanket discount, which trains your best guests to wait for money off. Use points to nudge and tiers to recognise. Insist that loyalty sees the whole bill, because spend is the truest measure of value. Use it to fill your quiet nights rather than discount them. And keep the data and consent on one guest record, run properly. Do that and loyalty becomes what it should be: the reason your regulars choose you again, not the reason they wait for a deal.
---
More in this series
Placeholder block. Link each as it publishes, per the editorial calendar.
FAQ
What is the best loyalty programme for a small restaurant?
The best one rewards real value across the whole bill, ties to a single guest record, and gives you control over who gets what, rather than offering a blanket discount to everyone. A scheme built into your till and CRM will almost always work better than a standalone app that only sees part of the picture. - q: "Do loyalty schemes actually bring people back?" a: >- A well-designed one does, because it gives a guest a reason to choose you over the alternative and a sense of being recognised. A badly designed one mostly gives money away to people who were coming anyway, or trains guests to wait for a discount. The design matters far more than the fact of having a scheme. - q: "Are points or tiers better for a hospitality loyalty programme?" a: >- They do different jobs. Points reward spend and can nudge behaviour, like filling a quiet night. Tiers recognise your most valuable guests and let you treat them differently. The strongest schemes use both: points to encourage, tiers to recognise. - q: "Why should loyalty connect to the till?" a: >- Because the till is where spend happens, and spend is the truest measure of value. A loyalty app that does not connect to the bill can only reward visits or guesses, not what someone actually spent. Loyalty that lives on the till rewards the right thing automatically. - q: "Can a loyalty scheme help fill quiet nights?" a: >- Yes, and this is one of its most underused jobs. Awarding bonus points or rewards for off-peak visits gives your regulars a reason to come on a Tuesday rather than a Saturday, shifting demand to where you have capacity rather than simply discounting the nights that were already full.