Five times the prize money bought nine per cent more entries.
That’s the Grant Burge AFL Grand Final promotion, which ran this winter through four retail channels as four separate campaigns. They sit side by side in the Trevor Services campaign file, which is the only reason the comparison is possible. BWS ran a $126,550 prize pool and closed on 1,775 entries. Liquorland ran $24,150 and closed on 1,623.
The on-premise version ran a $7,301 pool — under six per cent of what BWS put up — and closed on 1,144 entries. Roughly two thirds of the participation on a twentieth of the prize budget.
What the rest of the file looks like
We went back through every campaign on our platform at Trevor Services that has closed with entry data recorded: 57 of them, across cashback, sweepstakes, gift with purchase and standard entry mechanics. Taking the 38 closed campaigns that ran a genuine prize pool, the rank correlation between prize pool size and entry volume is 0.35. Positive, so the money isn’t doing nothing. Weak enough that if prize pool is the main number being argued over in the planning meeting, the meeting is arguing about the wrong number.
The clearest single case holds the brand constant. McGuigan ran two national promotions with almost identical prize pools — $40,900 and $43,050. One closed on 379 entries, the other on 267. Same brand, same order of prize money, a 42 per cent gap in participation. Whatever produced that difference, it wasn’t the size of the pool.
Then there is the top of the table. The three highest entry counts in our closed campaigns include an Electrolux gift-with-purchase offer that recorded 18,911 claims, a second Electrolux GWP on 3,917, and a Westinghouse offer on 3,535. All three carried a prize pool of zero.
That comparison isn’t like for like and we shouldn’t pretend it is. A GWP claim is someone collecting a guaranteed reward, not someone taking a ticket in a draw. But that’s the point rather than the caveat. Across our closed campaigns the median gift-with-purchase promotion recorded 1,288 claims; the median standard-entry promotion recorded 209 entries, and the median sweepstake 87. The mechanic that offers no chance of winning anything outperformed the mechanics built entirely on the chance of winning something, by roughly an order of magnitude.
Does a bigger prize pool get more entries?
A little, and rarely in proportion to what it costs. Across 38 closed Australian promotions with a prize pool, the rank correlation between pool size and entries was 0.35 — positive but weak, with a fifth of the prize money in one case delivering 91 per cent of the entries. Prize pool tends to set the ceiling on a promotion’s appeal rather than determine where within that ceiling it lands.
Two limits on that. This is our book, not the market: 57 closed campaigns weighted heavily towards liquor and whitegoods, which is what we run most of, and 38 is a small sample for a correlation. And channels aren’t comparable footprints — BWS has far more doors than an on-premise network, so part of that entry gap is distribution rather than persuasion. We’d argue that reinforces the finding rather than undermining it. If distribution is doing the heavy lifting, the prize pool isn’t.
What the research actually says — and what it doesn’t
The closest academic work to this question is about prize count rather than pool value, and it’s worth being precise about the difference before borrowing it. Research published in the Journal of the Academy of Marketing Science found that consumers deciding whether to enter a sweepstake largely can’t evaluate whether the number of prizes on offer is good or bad when they see it in isolation. Within a normal range, more prizes didn’t make people feel their odds had improved, and didn’t make them more likely to enter.
That’s a finding about how many prizes, not how much they’re worth. It doesn’t prove our point. What it does is describe the mechanism that would explain it: shoppers are poor at evaluating promotional magnitude without a reference point, so the number a brand agonises over is frequently a number the shopper never really reads.
It also sits awkwardly against one of our own rules of thumb. The Rule of Three in The Shelf Truth holds that one prize reads as impossible, three reads as possible and a hundred reads as probable. On this research, that reading only happens when the shopper has something to compare against — a competing promotion on the next shelf, a familiar category norm, a visual that makes the count concrete. Absent a reference point, the number is just a number. So the practical version of the Rule of Three isn’t “put more prizes in the pool.” It’s “make the odds legible.” Those are different jobs and only one of them costs prize money.
What the file says about the alternatives
Arguing that prize pool is weak is only half an argument. The obvious question is what’s stronger, and we should answer it with the same file rather than with assertions.
The clearest lever we can see is how often someone is allowed to enter. Closed campaigns that permitted more than one entry per day recorded a median of 864 entries. Campaigns capped at one entry per day recorded a median of 185. That’s six campaigns against seventeen, so treat it as a strong hint rather than a law — but it is a design decision that costs nothing in prize money and appears to move participation further than several multiples of prize budget did.
The more useful finding is one that went against us. We expected receipt requirements to suppress entries, because friction is supposed to be the expensive term in the 3-Second Equation. It doesn’t show up that way. Campaigns requiring a receipt recorded a median of 304 entries against 36.5 for those that didn’t, and holding the mechanic constant across standard-entry promotions the two are close to indistinguishable — 198 against 266. The likely explanation is selection rather than friction: the campaigns that ask for a receipt are the funded national ones with real media behind them, and that swamps the effect we were looking for. Campaign-level data can’t isolate form friction. You need funnel data — how many people started an entry and how many finished — and that’s a different measurement. We’d rather say that than repeat a rule of thumb our own numbers don’t support.
Where the pool number stops being free
There’s also a cost curve under the prize pool that is easy to miss while the figure is being set. In New South Wales, a trade promotion authority is only required once the total value of all prizes exceeds $10,000, with application fees from $488 for a one-year authority in 2026–27. South Australia, the ACT and the Northern Territory set their thresholds lower again, so a pool that clears one state’s line has usually cleared several. Crossing them is often the right call for a national campaign — it just ought to be a decision rather than something a brand backs into because a bigger number felt safer. The state-by-state detail is in the Trevor Services guide to promotional permits in Australia.
What to do with the money instead
None of this argues for a mean promotion. There is a floor, below which the reward isn’t worth the effort of claiming it, and shoppers find that floor quickly.
What the data argues against is reaching for the prize pool as the dial you turn when a promotion needs to perform better. In our file the campaigns that overperformed relative to budget were the ones offering a certain reward instead of a large one, allowing people to come back, or sitting in a channel that put them in front of more shoppers. Worth testing on the next brief, and we’re always happy to pressure-test a prize pool before it’s locked: if the pool were halved and the difference spent on distribution, entry frequency and making the odds believable, would the campaign do better or worse?
Our numbers suggest that for a lot of brands the honest answer is better, and that the question is almost never asked — because the prize pool is the easiest part of a promotion to argue about, and the hardest to be wrong about in public.
