Start with rewards actually redeemed
The cost of a restaurant loyalty reward depends on the offer, actual redemptions and purchases the offer might replace. Its menu value is not the whole calculation. An announced reward that nobody uses does not consume the same resources as a reward you serve.
The restaurant loyalty programme guide covers the broader setup. This article focuses on a budget you can check using your own costs, rather than a promised industry-wide return.
Worked example: a free item
Fictional example, in Canadian dollars: a drink sells for $5. Ingredients cost $1, packaging $0.30 and the assumed extra preparation cost is $0.40. Serving 40 rewards consumes 40 × $1.70 = $68 in variable resources.
That is not proof that the entire programme cost only $68. Include programme fees and administrative work. If a reward replaces a drink the guest would have purchased anyway, consider the revenue effect too. Menu value and variable cost answer different questions.
| Fictional example: a free drink | Cost |
|---|---|
| Ingredients | $1.00 |
| Packaging | $0.30 |
| Extra preparation | $0.40 |
| Total per reward | $1.70 |
| 40 rewards (40 × $1.70) | $68 |
Compare the same redemptions with a discount
Applying a $5 discount to 40 qualifying orders removes $200 of revenue, before considering any effect on visits or baskets. This does not show that free products are always better. Discounts can change what someone buys, and free items can displace an ordinary sale.
- Free item: 40 rewards, variable resources$68
- Discount of $5 on 40 orders, revenue removed$200
| What to record | Free item | Discount |
|---|---|---|
| Actual redemptions | Items served | Discounted orders |
| Immediate effect | Variable resources consumed | Revenue removed |
| Question to investigate | Did it replace a normal purchase? | Would the order have happened anyway? |
Toast's Canadian restaurant loyalty guide discusses margin when choosing offers. Your calculation still needs your own inputs; no return multiple is assumed here.
Do not mistake programme activity for gain
Keep membership counts, redeemed rewards, identified visits and sales for the period. Choose the comparison before the test: matched opening days, documented price changes and a note about seasonality. A larger average basket does not automatically offset fewer visits.
Another fictional scenario: you estimate $300 of additional contribution before rewards. Subtracting $68 of reward resources and $50 of fees leaves $182 in that scenario. If the additional $300 has not been demonstrated, the $182 is not a confirmed result.
- Assumed contribution$300
- Variable reward resources− $68
- Programme fees− $50
- Left in this scenario$182
Make the rule usable at the counter
Write down the eligible product, how it is earned and how redemption is validated. Train the team with a normal case and a duplicate case. Then follow a completed customer cohort instead of judging the programme by first-day signups.
Use the recipe and inventory food cost worksheets to check your inputs. Review the terms of use and discuss setup to agree on the scope before any commitment. Total member sales are not an automatic billing base.
