A/B testing gift card denominations for stronger campaigns
For B2B marketers running employee recognition or channel incentive programs, the dollar value printed on a gift card often shapes outcomes more than the brand stamped on the front. A $25 card can feel like a token gesture that gets forgotten in a wallet, while a $250 card can blow past the procurement threshold that made the program viable in the first place. Split testing is the cleanest way to settle the debate between low, mid and high values without relying on overseas benchmarks or the latest industry hearsay.
Australian programs bring a few local quirks worth weighing up before you start. The country runs on the Australian dollar and operates under GST rules that treat physical gift cards as a supply of goods at the point of redemption rather than at purchase, which can shift your reporting cycle. Consumer behaviour also skews heavily towards a small group of national retailers - Bunnings, JB Hi-Fi, Coles, Woolworths and Kmart dominate redemptions across most audiences, and each appeals to a slightly different demographic. Layer on seasonal swings like EOFY in June, the Melbourne Cup in early November, and the southern hemisphere summer Christmas, and you have a market where the "right" denomination can shift by the quarter.
That variability is exactly why a controlled A/B test earns its keep. Instead of asking stakeholders whether a $50 or $100 card will land better, you can run both against the same audience, measure what actually happens, and let the data decide. The work is in designing the test well, picking the metrics that matter locally, and reading the results with an Australian lens rather than copying a US case study.
Why denomination choice drives downstream behaviour
The number on the front of the card changes how recipients feel about the reward before they have even redeemed it. A $25 Bunnings card might be treated as a quick coffee top-up, while a $100 card gets earmarked for a tool purchase the recipient has been planning for months. That difference in perceived value feeds straight into redemption speed, basket size at the retailer, and the chance of the card sitting unused in a drawer until expiry.
Procurement teams in Australia tend to push back hardest on cards above $100, particularly for frontline reward programs where thousands of cards are issued each quarter. Marketers often settle on a safe middle value out of habit, but that approach can leave both engagement and budget efficiency on the table. A well-run test lets you back up a higher or lower value with evidence rather than anecdote.
| Test variant | Card value | Sample size | Redemption rate (90 days) | Average basket uplift | Cost per active recipient |
|---|---|---|---|---|---|
| Variant A | $50 | 1,200 | 71% | $58 | $0.42 |
| Variant B | $100 | 1,200 | 78% | $112 | $0.51 |
| Variant C | $150 | 1,200 | 74% | $148 | $0.69 |
Designing a clean split test for card values
A solid A/B test changes one variable at a time. In this case that variable is the denomination. Everything else - the retail partner, the card design, the delivery channel, the expiry window and the accompanying message - needs to stay identical across the test arms, or the results will be noisy and impossible to interpret.
Random assignment matters as much as the sample design. If you send the $50 cards to younger staff in NSW and the $100 cards to senior leaders in Victoria, you are not testing value, you are testing demographics. Use a randomised split inside your existing distribution channel, whether that is an HRIS export, a CRM segment, or a partner-supplied email list, so each audience member has an equal chance of landing in either group.
Keep these elements identical across every test arm:
- Retail partner (comparing Bunnings against Coles is a separate test)
- Card design, branding and packaging
- Expiry window and any T&Cs shown on the card
- Delivery channel such as email, SMS or printed mail
Choosing metrics that reflect Australian redemptions
Redemption rate is the obvious starting point, but it only tells half the story. A $50 card that gets redeemed 80% of the time but only adds a small amount to the average basket is a different commercial outcome to a $100 card redeemed 70% of the time that pulls a $140 basket through the door.
Track at least three signals side by side: redemption rate at 90, 180 and 365 days, average spend at the retailer per redemption, and the cost per engaged recipient. Add a short post-redemption survey through Qualtrics or even a basic Typeform to capture satisfaction scores, which are especially useful when pitching the result back to internal stakeholders in Sydney or Melbourne head offices.
Sample size and seasonal timing in the local market
Statistical significance does not care about your launch date, but your customers certainly do. Running a denomination test for two weeks in mid-November will catch the Melbourne Cup spending bump and skew redemptions towards hospitality and retail categories, while the same test in February will behave very differently.
Plan for at least four weeks of live data and a sample of 1,000 or more recipients per arm to get a reliable read on a national audience. If your program only covers a few hundred people, extend the test window or accept a wider confidence interval. Either way, avoid running the test across EOFY, Black Friday, or the pre-Christmas rush, or run separate analyses that strip those weeks out.
Splitting across physical and digital fulfilment
Australia has near-universal contactless payment acceptance at major retailers, which makes digital gift cards delivered by email or SMS a strong default for most B2B programs. They cost less to distribute, arrive instantly, and let you track redemption down to the hour.
That said, physical cards still earn their place for older demographics and for awards that carry ceremonial weight, such as long-service milestones. Test both delivery methods against the same denominations if budget allows, since the channel itself can change which value feels right. A $100 digital code can feel casual, while a $100 printed card from a recognised retailer can feel like a proper recognition moment.
Reading past the headline numbers
Averages hide the most interesting stories. Pull cohort splits by age band, state, and income tier before you call a winner. A $150 card might dominate in inner-Sydney professional services while a $50 card wins hands down among casual workforces in regional Queensland. That kind of segmentation often justifies keeping two denominations in market rather than forcing a single national value.
Watch for cannibalisation as well. If you already run a points-based recognition platform, an oversized gift card can pull engagement away from the everyday program rather than lifting overall participation. Look at total reward activity during the test window, not just gift card metrics, before declaring a winner.
Rolling out the winning value with confidence
Once the test points to a clear winner, phase the rollout rather than flipping the entire program overnight. Start with one business unit or one partner channel, confirm the real-world numbers match the test, then expand. This protects you if the test sample happened to over-index on a particular demographic.
Update finance, procurement, and any external fulfilment partners on the new standard value, and document the methodology so next year's campaign can build on it. A denomination test run once is useful; a denomination test repeated each year, with the previous year's results feeding the next hypothesis, becomes a genuine competitive advantage for an Australian incentive program.
Before locking in the new standard value, run through a short rollout checklist:
- Reconcile finance and procurement on the updated unit cost
- Brief the fulfilment partner on volume changes
- Confirm card inventory or digital code stock for the higher value
- Schedule a 90-day post-launch review to validate the test result