Skip to content
Fidella

Guide · 6 min

How to price a stamp card without losing money

Work from what the reward costs you to make, not what you sell it for. Then work out what that is as a share of what the customer spent getting there. Under about 5 cents in every dollar is normal, and cheaper than most advertising.

Most loyalty programmes are priced by looking at the venue down the road. That number encodes somebody else’s margins, ticket sizes and visit frequency.

Step one: what the reward actually costs you

Not the menu price. The input cost. A $5.50 flat white with $1.65 of beans, milk, cup and marginal labour costs you $1.65 to give away, not $5.50.

Getting this wrong in the pessimistic direction is the single most common reason an owner decides loyalty is too expensive.

Step two: what the customer spent to get there

Threshold times average ticket. Ten stamps at a $6 average is $60 of revenue.

Step three: express it as a percentage

$1.65 against $60 is 2.75 percent of revenue. That is your programme cost.

Compare it against what you pay for a customer through any other channel. It is difficult to lose that comparison, because most advertising buys attention rather than a repeat visit, and this only pays out when the visits actually happened.

The bands worth knowing

Under 3 percent is cheap. Three to 5 percent is normal and comfortable. Above about 8 percent you are either being very generous deliberately, in which case fine, or you have mispriced something.

Two adjustments most people miss

Not everyone finishes. Some proportion of issued stamps never reach a reward, which lowers the real cost. Do not build your plan on it. If a lot goes unused, it usually means people gave up. You have saved the money and lost the repeat visits you were paying for.

The reward is bought at your cost, not sold at your price. A customer redeeming a free coffee often buys something alongside it. That attach spend is real and it is why measured programme cost usually lands below the arithmetic.

The failure mode to avoid

Cheapening the reward to make the numbers work. A reward nobody wants costs you nothing and earns you nothing. If the maths is tight, raise the threshold and keep the reward worth having.

Common questions

Should I include labour in the cost?

Include the marginal labour only. If the reward is poured by someone already on shift, the incremental labour cost is close to zero, and loading full labour into it will talk you out of a programme that would have worked.

What if my margins are thin?

Raise the threshold rather than cheapening the reward. A reward nobody wants fails at any price.