Points based loyalty: the maths before the launch

A restaurant launches a points scheme: one point for every pound spent, and a free main course worth fifteen pounds for every hundred points. It sounds modest. Guests sign up, collect points and redeem them. A year later, the owner looks at the figures and realises the scheme has been giving away nearly a tenth of what its members spend, most of it to regulars who would have come anyway.

Points based loyalty can work, but only if the maths is done before the launch, not after. Three numbers decide almost everything: how fast guests earn points, what those points are worth when spent and how many are never spent at all. Get them right, and a scheme can reward regulars at a sustainable cost. Get them wrong, and it quietly eats several points of margin.

This guide works through the maths of a points scheme with example figures: the earn rate, the burn rate, breakage, how to cost rewards, what a scheme must achieve to pay for itself, and when points are not the right answer at all. Every venue and figure in it is an example.

A points scheme rewards guests for spending. Guests earn points on each visit, usually in proportion to what they spend, and exchange points for rewards: a free dish, a discount, a bottle of wine or an experience. The idea, common to every loyalty program, is that the prospect of a reward brings guests back more often than they would otherwise come.

Every scheme rests on the same few numbers. Change any one of them, and the cost of the scheme changes too. That is why the maths matters more than the design of the card or the app.

The earn rate is how many points a guest earns for each pound, or other unit, they spend. One point per pound is common because it is easy to understand. Some schemes give more points for certain dishes, days or times, to steer guests towards them.

On its own, the earn rate tells you nothing about cost. A scheme that gives ten points per pound may be cheaper than one that gives one point per pound, depending on what the points are worth. What matters is the earn rate together with the burn rate.

The burn rate is what points are worth when guests spend them. If a hundred points buy a reward worth fifteen pounds on the menu, each point is worth fifteen pence. With one point per pound spent, that is fifteen pence back for every pound: a reward rate of fifteen per cent of spend, at menu prices.

That figure, the face value returned as a share of spend, is the most important number in the scheme. In this example it is far too high for most restaurants. A reward rate of a few per cent of spend is more typical of schemes that last.

Breakage is the share of points that are earned but never spent. Guests forget about the scheme, move away, never reach the threshold or let points expire. Breakage reduces the scheme's real cost, because unspent points cost nothing.

If a scheme returns five per cent of spend in face value, and forty per cent of points are never redeemed, the face value actually redeemed is three per cent of members' spend. Breakage is real, but do not design a scheme that depends on guests failing to use it: guests notice, and it undermines the trust the scheme is meant to build.

Take an example scheme. Guests earn one point per pound and can exchange a hundred points for a reward worth five pounds on the menu: a reward rate of five per cent of spend. Expect around thirty per cent breakage, so seventy per cent of points are redeemed: a redeemed value of three and a half per cent of members' spend.

Now the cost depends on what the reward is. If it is a dessert the guest would never have bought, the cost to the venue is roughly the food cost of the dessert, perhaps a third of its menu price: a little over one per cent of members' spend. If it is a dish the guest would have paid for anyway, the cost is the full menu price: three and a half per cent.

The same scheme can cost one per cent or more than three per cent of members' spend, depending on what guests choose and what they would otherwise have bought.

How you cost a reward decides whether the scheme looks affordable. Costed at food cost, a free main course looks cheap: the ingredients might be a third of the price. Costed at menu price, it looks expensive. The truth lies in what the guest would otherwise have done.

If the reward brings a guest in on a quiet night, or adds a course they would not have ordered, food cost is close to the real cost. If it simply replaces something they would have paid for, the real cost is the lost sale. Our guide to loyalty cost: what a reward really takes out covers how to cost rewards honestly.

A points scheme pays for itself only if it changes behaviour: members visit more often, spend more per visit or choose the venue over competitors. If members would have come anyway, every point redeemed is a cost with no return.

Work out how much extra spending the scheme must create to cover its cost. If the scheme costs three per cent of members' spend, and the venue's margin on extra sales is, say, sixty per cent after food and variable costs, members must spend about five per cent more than they would have without the scheme just to break even. That is a real hurdle.

The hardest truth about points based loyalty is that the guests who earn the most points are often the regulars who would have come anyway. They visit weekly, love the venue and redeem rewards often. The scheme gives them money back without changing what they do.

That does not mean regulars should be excluded. But it does mean the scheme's value lies in turning occasional guests into regulars, not in rewarding the ones you already have. Look at whether the scheme moves occasional guests to come back sooner.

Loyalty redemption should be easy. A reward that is hard to claim, restricted to quiet days or hidden behind complicated rules frustrates guests and makes the scheme feel stingy. But easy redemption also raises cost, because more points are spent.

Balance the two openly. Set a reward rate the venue can afford with high redemption, rather than a generous headline rate that only works if guests struggle to redeem. Guests forgive a modest scheme that is easy to use far more readily than a generous one that is hard.

The number of points needed for the first reward shapes how guests feel about the scheme. A first reward within reach after two or three visits keeps guests engaged. One that needs fifteen visits feels unattainable, and many guests give up before they get there.

Thresholds also affect breakage. Distant thresholds increase breakage and reduce cost, but they also reduce the scheme's power to bring guests back. A near first reward, followed by larger ones further away, often works best.

Some schemes add loyalty tiers: guests who earn more reach higher levels, with better rewards or extra recognition. Tiers can motivate guests near the next level to visit a little more, and they let a venue recognise its best guests openly.

Tiers add cost if higher levels earn points faster or receive extra rewards. They can add value without much cost if the benefits are recognition: a preferred table, a welcome from the manager, an invitation to a tasting. Recognition is often valued more than its cost.

Points earned but not yet redeemed are a promise the venue has made. In accounting terms, they can be a liability: value the venue owes to guests. A scheme that has run for years may carry a large balance of unspent points.

Know how large that balance is and how it is likely to be redeemed. Ask your accountant how to treat it. A sudden wave of redemptions, perhaps when a scheme is changed or ended, can cost more than expected.

Points that expire after a period reduce the liability and encourage guests to visit, but expiry can feel unfair if it is not clear from the start. Changing the rules of an existing scheme, such as raising thresholds or shortening expiry, often angers the most loyal guests.

Set the rules carefully at launch, explain them plainly and change them only with notice and a clear reason. A scheme whose rules keep changing loses the trust it was built on.

Beyond the rewards themselves, a points scheme has running costs: software or cards, staff time to explain and process it, marketing to recruit members and the effort of maintaining the data. Some providers charge a monthly fee or a share of each redemption.

Add these to the cost of rewards when deciding whether the scheme is worth it. A scheme whose rewards cost two per cent of spend but whose software costs as much again is a four per cent scheme.

For many restaurants, points are not the best form of loyalty. A small venue that knows its regulars by name, a fine dining restaurant where guests come for occasions or a neighbourhood café with a loyal local crowd may do better with recognition, personal touches and occasional surprises than with a points scheme.

Our guide to twelve loyalty program ideas that are not points covers the alternatives, and our guide to a restaurant loyalty programme worth joining covers what makes any scheme work.

Before launching a points scheme, test the maths with your own figures: your average spend, how often your guests visit, your food cost and your margin. Work out the reward rate, expected breakage and the extra spending the scheme must create to break even.

If you can, trial the scheme with a small group of guests for a few months and measure what changes: visit frequency, spend per visit and how often rewards are redeemed. A trial costs little and can save a scheme that would have cost a great deal.

After launch, measure the scheme every month: how many members, how often they visit compared with non-members, how much they spend, how many points are earned and redeemed and the balance outstanding.

Compare members with similar guests who did not join, and compare members' behaviour before and after joining. If members visit no more often than they did before, the scheme is a cost without a return, and it is time to rethink it.

Not every pound of spend is equal. A scheme that earns points on drinks with a high margin, or on quieter days, steers guests towards the spending that helps the venue most. A scheme that earns the same on everything treats a busy Saturday main course the same as a quiet Tuesday dessert.

Some schemes offer extra points at quiet times or on particular dishes. Use this carefully: it can shift behaviour, but complicated rules confuse guests and staff. One or two simple bonuses are usually enough.

Staff are the face of any points scheme. They explain it, sign guests up, answer questions and process rewards. A scheme the team does not understand, or finds awkward, will be explained badly and used less.

Train the team on how the scheme works, why it exists and how to explain it in a sentence. Make processing rewards quick on the till. A scheme that slows down the end of the meal will be quietly avoided by busy servers.

Points have value, and anything with value can be misused: points added to a friend's account, rewards redeemed without a real visit, accounts shared between many people. Most misuse is small, but it adds up.

Record who adds points and processes rewards, review unusual patterns and make sure points can only be earned on real, paid visits. A few simple controls stop most misuse without making the scheme harder for honest guests.

Sometimes a points scheme needs to end or change, because it costs too much, nobody uses it or the venue wants a different approach. Plan it carefully. Give members notice, let them use their points for a reasonable time and explain what will replace the scheme.

Ending a scheme badly, with points cancelled overnight, damages trust with the very guests the scheme was meant to keep. Ending it well, with time to redeem and something better to replace it, can strengthen it.

Decide early what counts towards points, because it changes the maths. Most schemes exclude tips and service charges, which belong to the team, and gift card purchases, which would otherwise earn points twice: once when the card is bought and again when it is spent.

Other choices are less obvious. Should points be earned on delivery orders, on deposits, on drinks at the bar before a meal, on private hire? Each one you include raises the cost of the scheme, and each one you exclude needs explaining. Write the rules down in plain language, so that staff and guests read them the same way.

A guest who is told at the till that their bar drinks do not count, after assuming they would, feels cheated. Clear rules at sign up prevent the argument at the bill.

Guests rarely do sums at the table. If a hundred points are worth five pounds and a pound earns ten points, many guests will never work out what their visit earned them. A scheme whose value is hard to grasp is a scheme that motivates nobody.

Keep the arithmetic simple enough to say in a sentence: one point per pound, and a reward at a round number. Show the balance and the next reward together, so the guest sees how close they are. Avoid fractions of points and rules that change by day or dish, unless the scheme genuinely needs them.

The test is whether a new server can explain the scheme correctly after one shift. If they cannot, guests will not understand it either.

An invented brasserie plans a scheme with one point per pound and a fifteen-pound reward for every hundred points: fifteen per cent of spend back at menu prices. The owner works the numbers through before launching. With thirty per cent breakage, the scheme would return about ten per cent of members' spend, mostly to regulars who already visit weekly.

The owner redesigns it: one point per pound, with a five-pound dessert or drink for every hundred points, and a first reward after fifty points to bring guests back sooner. The reward rate falls to five per cent at menu prices, and well under two per cent once breakage and food cost are counted. Higher tiers bring recognition rather than more points.

After six months, members visit more often than they did before joining, the scheme costs less than two per cent of members' spend and the brasserie can see the outstanding balance on its books each month.

servearound does not run a points scheme, a stamp card or a loyalty app, so it does not track points, rewards or redemptions. What it provides is recognition built from the service. Every booking, order and check adds to the guest's profile, and guest tiers, new, regular, VIP and VVIP, are worked out from visits and spend over a window each venue sets, with a tier set by hand never overwritten.

Guest analytics show how many guests came back, the repeat rate, the tier mix and how many return from each week's first visits, which is the evidence a loyalty idea must move. With Campaigns, guests who opted in can be invited by tier, visits or spend to a tasting or an early look at a new menu. See the guests module and Analytics.

A scheme where guests earn points as they spend and exchange them for rewards.

How many points a guest earns for each pound they spend.

What points are worth when spent. Together with the earn rate, it sets the share of spend returned.

The share of points never redeemed. It lowers cost, but a scheme should not depend on it.

By what the guest would otherwise have done: food cost if the reward is extra, menu price if it replaces a paid purchase.

Usually a few per cent of spend, not the double figures many first designs offer.

Only if higher tiers earn more or receive extra rewards. Recognition costs little.

When the venue knows its regulars well and recognition would do more than rewards.

Points based loyalty stands or falls on its maths. The earn rate and the burn rate together set the share of spend returned to members; breakage reduces it; and the real cost of each reward depends on whether it adds to what the guest buys or replaces it. Many first designs return far more than the margin can support.

Work the numbers through with your own figures before launching, cost rewards honestly, aim for a modest reward rate that works even with high redemption, keep the first reward within reach, use tiers for recognition, track the outstanding balance and measure whether members actually come back more often.