Customer loyalty programs: the time-to-first-reward test
In short
A customer loyalty program only changes behaviour if the first reward arrives before the customer decides whether to return. Divide the smallest reward by order value times earn rate, then multiply by the gap between orders. If that takes longer than one natural purchase cycle, the programme pays customers who were coming back anyway.
What is a customer loyalty program for?
A customer loyalty program exists to change one behaviour: getting a customer who would not otherwise have come back to place another order. Everything else — the app, the tiers, the badge — is delivery. If you want the definitions and the four standard types, the loyalty program guide covers them along with what each one costs.
This page is about the step before any of that: whether a programme can reach your customers at all. Most fail on that question, quietly, and no dashboard reports it.
The test most customer loyalty schemes fail
Before choosing mechanics, work out how long a typical customer waits for their first reward. A programme only changes behaviour if the reward is close enough to be felt. If the first reward arrives after the customer has already decided whether to come back, it pays out to people who returned anyway.
The calculation needs four inputs you already have: average order value, earn rate, the value of the smallest reward, and the gap between orders.
Orders to first reward = smallest reward ÷ (order value × earn rate), rounded up. Then months to first reward = (orders − 1) × months between orders, because the first order starts the clock.
| Store | Order value | Earn rate | Smallest reward | Points value per order | Orders to first reward | Months between orders | Months to first reward |
|---|---|---|---|---|---|---|---|
| Coffee | 50 | 5% | 10 | 2.50 | 4 | 1 | 3 |
| Skincare | 200 | 5% | 25 | 10 | 3 | 6 | 12 |
| Furniture | 400 | 5% | 50 | 20 | 3 | 12 | 24 |
The same earn rate produces three different programmes. The coffee customer sees a reward inside a quarter, while the furniture customer waits 24 months, by which point the programme has had no chance to influence the one decision that mattered.
The skincare row is the dangerous one, because it looks reasonable on a planning sheet. At 12 months to first reward, any points that expire after a year are lost at the exact moment they would have become useful, and a customer who notices that will read the programme as a trick.
How to design a customer loyalty plan in five decisions
Take these in order. Each one can end the process, and that is the point of doing them first.
- Is your purchase cycle short enough? Run the time-to-first-reward test. If the answer is longer than a customer's natural gap between orders, a points programme is the wrong tool; look at a paid programme or at the second-order experience instead.
- What is the reason people do not come back? A programme addresses forgetting and indifference. It does not address a disappointing product or a slow delivery, and it will not tell you which one you have.
- What can you afford? Set the ceiling from contribution margin, not from competitors. The break-even arithmetic sets how much extra spend the programme must produce to pay for itself.
- Which mechanic? Only now. Choose the one that makes the first reward reachable inside one natural purchase gap.
- How will you know? Decide before launch which customers will not be offered it. Without that group there is no answer later, only a members-versus-non-members chart that says what it would have said anyway.
The order matters more than the choices. Stores that start at step four — usually because a platform demo started there — arrive at step five with no way back.
Client loyalty programs in B2B: why points rarely fit
In B2B the buyer and the payer are usually different people. A points balance accrues to a company account, where nobody feels it, or to an individual buyer, where it becomes an incentive the client's finance team did not agree to. Neither changes the decision the programme was meant to change.
What does work in B2B is structural rather than promotional: volume terms written into the contract, service levels that improve with tenure, and early access that matters operationally. These are loyalty programmes in effect, but they are priced into the relationship rather than bolted on as a points layer, and they are negotiated rather than displayed.
If a client loyalty programme is on the table, ask who at the client would ever see the reward. If the honest answer is nobody with influence over the renewal, the budget is better spent on the account itself.
What to measure from the first day
Measure three things from launch, not after the first review meeting:
- Time to first reward, actual. The planning figure is a guess. The real median, from order history, tells you whether the design works for the customers you actually have.
- Repeat rate for a holdout group. A random slice that is never offered the programme. It is the only comparison that isolates what the programme did.
- Cost as a share of revenue, including unredeemed points. Outstanding balances are a liability, and they are redeemed in bursts that land in quarters that did not earn them.
Repeat rate on its own will mislead you here, for the reason set out in repeat purchase rate vs retention rate: it moves with your acquisition mix, and a programme launched in a growth quarter will look like it failed. Read it against cohorts, and against the holdout, before concluding anything about the programme or customer retention more widely.
Common questions
What is a good time to first reward?
Shorter than the natural gap between two of your customers' orders. If a customer typically reorders every two months, they should be able to see a reward by their second or third order. Any longer and the reward arrives after the return decision it was meant to influence.
Should loyalty points expire?
Expiry limits your liability, but it must be longer than time to first reward, or points vanish just as they become usable. Set expiry from your purchase cycle, and warn customers before it happens rather than letting balances disappear silently.
Is a paid loyalty program better than points?
For long purchase cycles, often yes. A fee selects for customers who already intend to return and makes the programme self-funding, whereas points spread a cost across everyone, including people who would never have reached a reward.
Do small stores need a customer loyalty program?
Not by default. A small store with a short purchase cycle can test the idea with a single well-timed offer before committing to a platform and a permanent cost. Run the time-to-first-reward test first; it is free and it often ends the discussion.
See it on your data
On a twenty-minute call we connect one ad account and one month of your orders, and work out the gap between what the platform claims and what your shop took. Profit and POAS stay blank until you enter your margin and costs per order.
Book a demo