Gamification Reward Design: Discounts or Chance-Based Prizes?
By Haroon Haneef · Published · 8 min read
On this page
- Decide what happens after the reward
- Separate a certain benefit from a certain value
- What does the research actually support?
- Asda shows how a variable reward can still be certain
- Hypothetical example: equal face value, different exposure
- Make checkout honour the reward screen
- Test the promise through to purchase
- Frequently asked questions
- Bring a reward brief, not just a prize idea
Choose gamification reward design backwards from redemption: decide what the customer should do, what benefit they can actually use, and what the business can afford. Compare a known discount, a variable reward everyone eligible receives, and a prize some players will miss. Test purchases and contribution after reward costs before declaring a winner.
Decide what happens after the reward
Start with a completed customer action, not a prize catalogue. Write the redemption route before choosing the game: who qualifies, what they receive, where they use it and who resolves a failed claim.
For a purchase campaign, fill that blank with a valid offer redemption on a relevant order. For a discovery campaign, it might be viewing a recommended product. If the useful result is a recommendation, consider whether a monetary reward is needed at all.
Keep the wider campaign planning in the gamified marketing guide. The decision here is narrower: what promise should the reward make?
Separate a certain benefit from a certain value
Use these as working definitions for your campaign brief. They describe reward allocation after the participant meets the disclosed conditions; any skill requirement belongs in a separate field.
| Reward structure | Promise to an eligible participant | Recommended use | Decision to settle before launch |
|---|---|---|---|
| Fixed benefit | A known coupon or benefit | A straightforward offer where clarity is the priority | Can the margin support every valid redemption? |
| Variable benefit for everyone eligible | A benefit is certain; its value is revealed later | A reveal where nobody eligible leaves empty-handed | Is the smallest outcome worth the requested effort? |
| Chance-only prize | Some outcomes give no benefit | A campaign where the prize opportunity is the proposition | Are the no-win outcome and claim process clear before play? |
A mystery discount and a chance to win are different promises. Write the actual outcome structure in the brief rather than leaving “random reward” for the developer to interpret.
GameOn Brands lists Scratch Card, Quiz and Spin The Wheel among its formats. Choose the reward promise first, then use the ecommerce gamification ideas guide to choose an interaction that fits it.
What does the research actually support?
The evidence supports testing reward design; it does not name a universal winner.
Martin Adam, Annika Reinelt and Konstantin Roethke compared monetary reward designs in registration systems. Their field and online experiments found higher registration likelihood for gamified rewards than for a non-gamified certain reward, with chance-based gamified rewards highest in the tested designs. Their distinction includes uncertainty about reward value after successful game performance; it should not be read as proof that a no-win draw is superior. The measured outcome was registration, not profitable purchases (original paper).
Celia Gaertig and Joseph P Simmons found that probabilistic promotions outperformed equivalent sure discounts only when the sure discount was or seemed trivial. Their studies concern promotion preferences, rather than an Indian branded-game profit benchmark (original paper).
My recommendation: start with a fixed benefit as a clear baseline, then test whether uncertainty earns its operational complexity. If the fixed offer is already meaningful, do not assume a larger headline prize will improve the purchase decision. A registration result is a reason to investigate, not a sales forecast.
Asda shows how a variable reward can still be certain
Asda’s Rewards app promotion offered a Cashpot bonus on every qualifying spin: customers spent at least £5, scanned their app and could receive 50p, £1, £5 or £100 during its September 2023 promotion. The brand’s announcement describes the design, without establishing a causal performance result (Asda announcement).
The useful design decision is the separation of eligibility from allocation: qualifying purchase first, variable benefit afterwards. For your brief, ask whether the reward should encourage the current order or give an existing customer a reason to return. Do not copy the prize ladder without also deciding that sequence.
RMH’s own documented GameOn reward history includes coupon discounts from 5% to 50%, with occasional free products. That is real reward experience, but the available record does not establish allocation odds, an everyone-wins policy or comparative profitability. It cannot tell you which structure will win your test.
Hypothetical example: equal face value, different exposure
This is an illustrative calculation, not a real RMH campaign or result. Assume 1,000 eligible completions and fixed-value coupons in rupees. All figures below are assumptions.
| Structure | Assumed allocation | Expected face value per completion | Expected issued face value |
|---|---|---|---|
| Fixed | Everyone receives ₹50 | ₹50 | ₹50,000 |
| Variable, everyone receives something | 50% receive ₹20; 50% receive ₹80 | ₹50 | ₹50,000 |
| Chance-only | 10% receive ₹500; 90% receive nothing | ₹50 | ₹50,000 |
Each allocation totals 100%. The variable calculation is 0.5 × ₹20 + 0.5 × ₹80 = ₹50; the chance-only calculation is 0.1 × ₹500 + 0.9 × ₹0 = ₹50.
Equal expected face value is not equal actual cost. If exactly 200 fixed coupons are redeemed in full, the discount given is ₹10,000. Under the other structures, calculate cost separately for each reward tier using its actual redemptions. Do not apply one assumed redemption rate to every tier and call it measured economics.
Random allocation also makes expected liability different from a hard cap. With independent chance-only draws, all 1,000 outcomes could theoretically receive ₹500, creating ₹500,000 in issued face value. A finite inventory would control that exposure differently; its allocation and exhaustion rules would need to match the participant-facing promise.
For percentage coupons, model eligible basket values and any caps. For free products, include product, packing and delivery costs. Then compare contribution after discounts, fulfilment, campaign costs, returns and support against the control. Equal prize budgets alone do not make two offers commercially equivalent.
Make checkout honour the reward screen
Treat the result screen and checkout as a single promise. On Shopify, amount-off discounts have separate settings for eligible products, minimum purchase requirements, customer eligibility, usage limits, combinations and active dates (Shopify documentation).
Use this launch checklist as an operating recommendation:
- State qualifying actions and possible outcomes before play, including a no-benefit outcome.
- Show eligible products, qualifying spend, expiry and combination restrictions before the participant commits.
- Assign an owner for stock, coupon delivery, missing rewards and disputed claims.
- Test repeat visits, interrupted sessions, exhausted allocations and the closed-campaign screen.
- Preserve a valid earned reward through a retry; decide how duplicate claims are handled.
- Provide a readable result and claim route without making an animation the only way to understand the outcome.
For claim forms, W3C guidance calls for automatically detected input errors to identify the affected item and describe the error in text (error-identification guidance). Apply that to an invalid field rather than silently reloading the form. “Something went wrong” is a support ticket wearing a sentence.
For an Indian prize campaign, have a qualified professional check the final structure and terms. This guide is general information, not legal advice.
Test the promise through to purchase
Use a controlled test plan agreed before distribution. The following worksheet is a recommendation, not a claim that any reward will improve performance.
| Brief field | What to write |
|---|---|
| Audience and allocation | Comparable eligible visitors assigned to each treatment; define repeat-visitor handling |
| Comparison | Fixed offer versus variable or chance-only offer; include a no-game baseline where feasible |
| Conditions held steady | Placement, audience, products, claim route and timing |
| Economic assumptions | Expected reward value, maximum exposure and redemption assumptions by tier |
| Primary outcome | Contribution per eligible visitor after discounts and relevant campaign costs |
| Diagnostic events | Exposure, start, completion, reward issued, claim attempt, redemption and purchase |
| Guardrails | Margin, failed claims, support demand, cancellations and returns |
| Test boundary | Start, end, sample plan and decision rule agreed before results arrive |
Compare the full assigned audiences, including people who never start or finish. Looking only at players who redeem leaves out the people the treatment failed to move.
If the goal is retention, add a defined follow-up window for repeat purchase after the incentive ends. Keep registration, redemption and repeat purchase as separate outcomes; success at one does not answer the others.
Frequently asked questions
Should every player receive a reward?
I recommend offering every eligible participant a useful benefit when the campaign promise is an offer reveal. Chance-only prizes suit a different promise. Asda provides a documented example of variable bonuses without an empty qualifying spin (campaign announcement).
Is a mystery discount the same as a prize draw?
No, under this guide’s definitions: a mystery discount promises a benefit whose value is revealed later; a chance-only prize includes a no-benefit outcome. Asda’s varying Cashpot bonuses illustrate the former distinction (source).
Are chance-based rewards better than fixed discounts?
There is no universal winner in the cited evidence. Research on uncertain price promotions found their advantage depended on whether the equivalent sure discount seemed trivial (paper). Test your offer against a clear fixed baseline and judge the commercial outcome.
What should be visible before someone plays?
Show the qualifying action, possible outcomes, restrictions, expiry and claim route before commitment. Check those promises against the actual commerce settings; Shopify documents eligibility, minimum purchase, usage, combination and date controls separately (documentation).
Bring a reward brief, not just a prize idea
Bring your campaign goal, audience, reward budget, maximum exposure, redemption restrictions and test metric to RMH’s campaign team. Use those decisions to discuss a suitable GameOn Brands campaign, with a reward the customer can understand and the business can deliver.