When I help SMEs design first-purchase incentives, the first question I always ask is: “How big can the reward be before it actually destroys margin?” It’s an important, simple-sounding question — and one that too many programs answer with gut instinct rather than a repeatable formula. Below I share the exact framework I use to calculate a defensible reward ceiling for first-time shoppers, with the variables you need, a worked example and a small scenario table so you can plug in your own numbers.
Why you need a reward ceiling (and what it actually means)
A reward ceiling is the maximum value of an incentive you can offer to a first-time shopper without delivering negative contribution to your business once you account for the cost of the reward and the incremental revenue it generates. Put differently: it’s the highest reward that still leaves you with non-negative incremental gross profit after acquisition and redemption costs.
Setting a ceiling protects margin and forces you to design incentives that are strategic (drive repeat purchase) rather than purely promotional (give away margin for no long-term value).
Key variables you must estimate
- AOV — Average Order Value of that first purchase or the expected next purchase the reward applies to.
- Gross margin (GM%) — Your product gross margin percentage (revenue minus COGS, before marketing and fulfilment).
- Baseline repeat purchase probability — Probability the customer would return without any reward.
- Post-incentive repeat probability — Probability the customer will return because of the incentive.
- Incremental orders (∆Orders) — Expected additional number of orders attributable to the reward within your chosen measurement window.
- Redemption rate — Percentage of issued rewards actually redeemed (affects cost).
- Customer acquisition cost (CAC) — The incremental marketing spend to acquire that first-time shopper (if you’re only interested in the reward part, you can treat CAC separately — but to understand full economics include it).
- Reward cost model — Is the reward a fixed voucher (£10 off) or a % discount? If % we model as a margin reduction rather than cash paid out.
- Time horizon — The period over which you measure uplift (30/90/365 days). Longer windows capture more LTV but increase uncertainty.
The simple formula I use
The reward ceiling is the value of the reward (R) such that:
Incremental Gross Profit from additional orders ≥ Reward cost (net of any margin retention) + Allocated CAC
Expressed algebraically, for a fixed-reward voucher applied to one future order:
| R | = | (AOV × GM% × ∆Orders) − CAC × (1 / Redemption rate) |
Notes on the formula:
- If CAC is already sunk (you measure reward cost only), set CAC = 0.
- If the reward is a percentage discount, model its effect as reducing GM% on the redeemed order rather than as a cash payout — but the same logic applies.
- ∆Orders is the incremental number of orders caused by the reward: ∆Orders = Post-incentive repeat probability − Baseline repeat probability (over your time horizon).
- Redemption rate stretches the gross reward issued to the expected cost: if only 50% redeem, you can issue twice the face-value and still pay the same actual cost. But be careful: a very low redemption rate often signals poor mechanics and poor customer experience.
A worked example
Say you sell homeware online. Your inputs:
- AOV = £60
- Gross margin = 55% (so gross profit per order = £33)
- Baseline repeat probability (30 days) = 8%
- Post-incentive repeat probability (30 days) = 20% → ∆Orders = 12% = 0.12 orders per customer
- Redemption rate = 60%
- CAC allocated to this acquisition = £8
Incremental gross profit = AOV × GM% × ∆Orders = £60 × 0.55 × 0.12 = £3.96
Reward ceiling (face value you can issue) = (Incremental gross profit − CAC) ÷ Redemption rate
So R = (£3.96 − £8.00) ÷ 0.60 = −£6.73
That negative result is the crucial insight: with these inputs, the expected incremental gross profit in the chosen window (30 days) does not cover CAC, so you cannot sustainably issue a positive-value reward without losing money. This tells me either:
- Increase the time horizon (capture more future orders),
- Improve the post-incentive repeat probability (make the offer more compelling),
- Decrease CAC,
- Or accept the loss as a calculated longer-term investment if you can prove longer-term LTV makes up the gap.
Scenario table — quick reference
| Scenario | AOV | GM% | ∆Orders | Redemption | CAC | Max reward (R) |
|---|---|---|---|---|---|---|
| Conservative | £40 | 45% | 0.05 | 50% | £6 | (£40×0.45×0.05 −6)/0.5 = −£19.00 |
| Realistic | £60 | 55% | 0.12 | 60% | £8 | −£6.73 (see example) |
| Optimistic | £70 | 60% | 0.25 | 70% | £6 | (£70×0.6×0.25 −6)/0.7 = £10.29 |
This table highlights how sensitive the ceiling is to ∆Orders, AOV and GM%. Small increases in repeat rate or AOV can move a campaign from loss-making to profitable quickly.
Practical tips to raise your ceiling (i.e., extract more margin-friendly value)
- Target high-AOV segments: first-time shoppers who buy premium items give you more margin headroom.
- Design rewards to drive profitable behaviour: make rewards apply to specific categories with higher margin, or set minimum spend thresholds (e.g., £10 off £50+).
- Use time-limited but not immediate discounts: a “next order” credit that expires in 60–90 days increases chances of additional full-price purchases before redemption.
- Reduce CAC: test cheaper acquisition channels or optimise paid ads to lower the allocated cost for these cohorts.
- Increase perceived value without matching face-value: offer tiered rewards, early access, or samples — sometimes perceived value drives repeat without full cash cost.
- Measure and iterate: A/B test your offers and feed real redemption and repeat data back into the model rather than relying on guesses.
What I track in the first 90 days
When I launch a first-purchase incentive, I track these KPIs weekly and feed them into the ceiling model:
- Acquisition CAC by channel
- Reward issuance and redemption rate
- Post-purchase repeat probability (30/60/90 days)
- Average order value on redeemed orders vs non-redeemed orders
- Incremental gross profit per cohort
These numbers quickly show whether your assumptions are realistic and where to prioritise optimisation (channel, creative, reward mechanics).
If you’d like, I can help you build this calculation in a small Excel template with your own inputs so you can test scenarios and see a recommended reward ceiling per campaign. It’s the easiest way to stop guessing and start offering incentives that actually scale retention — not just discounts that erode margin.
