Almost every promotional brief that reaches us has a duration in it, and almost every one of them is a single number. Eight weeks. Six weeks. “Runs through spring.” The number is usually inherited rather than decided — it’s how long the feature is booked, or how long the media flight runs, or how long the display stays up.

Here’s the problem with one number. In our own campaign records, the gap between the last day a purchase qualifies and the last day a customer can claim is one day for prize draws and 92 days for cashbacks. Same brief format, same planning meeting, three months apart. A promotion doesn’t have one length.

A promotion has three clocks, not one

The first is the sell period — the window in which a purchase qualifies. This is the one everybody sets, because it’s the one the retailer and the media plan care about.

The second is the claim or entry window — how long a buyer has to actually do the thing: enter the draw, upload the receipt, submit the cashback.

The third is the fulfilment tail — the time between a valid claim and the money or the prize reaching the person. Verification, draw, winner contact, payment run, dispatch. It lives entirely on the operational side, which is why most briefs don’t mention it at all.

Set only the first clock and the other two default to whatever the platform, the terms template or the finance calendar happens to do.

How long should a promotion run?

For entry mechanics — prize draws, instant wins, sweepstakes — the entry window should close on the last day of sale, with a sell period of around six to nine weeks. For cashback, the claim window should stay open roughly 90 days after the last day of sale. Gift with purchase splits into two different shapes and needs a decision rather than a default.

The numbers behind that

Below is the full sample from the Trevor Services campaign book: every promotion we’ve delivered since 2019 that recorded both a final sale date and a final claim date. 54 campaigns. Nothing excluded.

Three things in that table are worth saying plainly, including the parts that don’t flatter it.

The entry-mechanic result is the solid one. Across 37 draw and sweepstake campaigns the median gap is a single day. Entry closes when the sell period closes, consistently, and the wide range comes from a handful of multi-phase promotions where one set of dates covered several draws.

The cashback result is real but the sample is small. Nine campaigns, and the middle of the distribution is tight — 90, 92, 92, 92, 122 days — with one campaign at a single day and two long-running programmes at 1,849 and 2,244 days dragging the top. Nine is enough to notice a convention. It is not enough to call it a law, and we’d rather say so than round it into one.

The gift-with-purchase number is not a recommendation and shouldn’t be read as one. That median of 761 days is an artefact of a genuinely bimodal set: four campaign-shaped promotions at −12, 1, 61 and 92 days, and four always-on offers running past four years. There is no typical GWP claim window in our book, because GWP is doing two different jobs. The useful question isn’t “how long” — it’s which of the two you’re actually running.

And the obvious caveat: this is our book, not the market’s. These are campaigns Trevor Services scoped and built, so the conventions in it are partly our own. Take the entry-versus-redemption contrast as the finding, and the specific day counts as a starting point to argue with.

Why the wrong calendar gets used

Nearly everyone’s instinct about promotional timing was formed by prize draws, because prize draws are nearly all anyone runs.

In the live Australian promotions Trevor Services tracks, prize draws and instant wins account for 123 of 181 campaigns currently in market. Cashback accounts for four. If your mental model of “how long a promotion runs” was built on that distribution, it was built on the mechanic where entry closes on the day — and it will be wrong, by about three months, the first time you apply it to a cashback.

The structural reason is simple. A prize draw closes with an event. There’s a draw date, and everything before it is entry accumulation, collected at or near the moment of purchase. Adding weeks doesn’t make it work harder — past a point it just spends display time and media weight to collect a thinner stream of entries. If entry volume is the problem, length is rarely the fix; friction usually is.

A cashback doesn’t close with an event. It closes with the last person who bothers. The buyer purchases, gets the product home, finds the receipt, and claims — and those three steps are separated by ordinary life. The 90-odd day convention isn’t generosity. It’s roughly how long it takes a normal household to get around to it.

Shortening the claim window is a price cut you didn’t approve

Compress that window and you don’t get a faster campaign, you get a cheaper one, because more people miss the deadline. That gap between purchases and claims is slippage, and it’s a legitimate part of how cashback economics work.

But there’s a difference between planning for it and pocketing it. If you’re tightening the claim window because you want the redemption rate down, model it, price it, and put the assumption in the business case where someone can argue with it. If you’re tightening it because the promotion “ends on the 30th” and nobody thought about it, you’re taking the same commercial benefit by accident — and paying for it in escalations, complaints and manual goodwill payments that land on a team who never saw the calendar.

A cashback with a 30-day claim window is a different offer to the same cashback with 90 days. It should be signed off as one.

The pack outlives the promotion

If the offer is printed on the pack, the pack becomes a piece of advertising whose retirement date you don’t control.

The ACCC uses precisely this scenario as a worked example. It describes cans of deodorant shrink-wrapped with “$3 Cash Back” where the offer had expired a week earlier, and the expiry could only be seen in the fine print after the packaging was opened. The ACCC’s guidance on cash back offers, gifts and prizes is that the packaging is misleading, because the bold representation was made without clear mention of the limitations.

That’s a duration problem wearing a compliance costume. Stock doesn’t clear when the campaign ends. On-pack offers keep selling themselves from pantries, warehouses and the back of the shelf long after the media stops — which is the argument for treating 90 days as a floor rather than a ceiling, and for checking how long the point-of-purchase display stays up relative to the offer printed on it.

The fulfilment tail carries its own obligation. The same ACCC guidance makes it unlawful to offer a prize or gift and then fail to provide it as offered, or fail to provide it within the time specified — or, where no time is specified, within a reasonable time. An unstated fulfilment tail isn’t a neutral omission. It hands someone else the job of deciding what “reasonable” means.

What the permit calendar does to your start date

Duration has a hard floor at the front as well, and it’s the one that most often surprises people.

In New South Wales, an authority is required when the total prize value for a single trade promotion exceeds $10,000. Where an authority applies, NSW Fair Trading requires a copy of the gaming rules at least 10 working days before the promotion takes place, and the activity cannot commence until that notification has been given. Two working weeks, sitting in front of your start date, before anything goes live. The rules sit under the Community Gaming Act 2018.

The back end is regulated too. Under the same NSW guidance, if the rules don’t state a timeframe for an activity requiring an authority, the operator must keep an unclaimed prize for at least three months before a new winner can be drawn. Your promotion has a tail whether or not you wrote one. The only choice is whether you set it or inherit it.

Thresholds and processes differ across the states, which is a separate planning exercise — we’ve covered the detail in our guide to competition permits in Australia.

Three dates, set on purpose

Pick the mechanic, then let the mechanic set the calendar. Close entry on the last day of sale for a draw. Hold a cashback open about 90 days past it, longer if the offer is on-pack. Decide which kind of gift with purchase you’re running before you date it at all. Then write the fulfilment tail into the terms as a stated number of days, because it exists whether or not you name it.

The claim window is the only one of the three that is simultaneously a customer-experience decision, a compliance position and a line in the budget. It is usually the one nobody owns. None of this costs anything while it’s still a date in a planning document, and all of it is expensive afterwards, because by then the packs are printed.

If you’d like a second opinion on your dates before that, talk to us.

Three campaigns. Identical $20,000 prize pools. They finished on 52 entries, 67 entries and 3,768 entries.

All three ran on the Trevor Services Salesforce platform for Australian brands — an appliance instant win, a wine promotion sold through a liquor wholesaler’s trade base, and a consumer wine campaign running through retail. Same prize money, a seventy-fold difference in entries. That gap is worth holding onto the next time a budget meeting opens, as they nearly always do, with the question of whether the prize is big enough.

What a promotion prize is worth on the open market

Before arguing about $20,000 versus $50,000, it’s useful to know what everyone else is spending. Of the 181 live Australian promotions Trevor Services currently tracks, 141 carried a stated prize value. Across all mechanics the median was $19,000, but that figure mixes formats that aren’t comparable. The number to use for a conventional single-winner prize draw is around $14,000.

The huge totals in the market are almost all instant wins, and they’re a different purchase entirely — the largest in the set carried a pool above $5.7 million running through licensed venues, spread over thousands of small prizes. That’s buying frequency of winning, not size of win, and it belongs in a different line of the budget.

So $14,000 or so is where the Australian market sits for a national draw. Going well above it is a decision that needs a reason, and “the prize felt small” isn’t one.

Does a bigger prize get more entries?

Not much, on our numbers. We hold both the total prize pool and the final entry count for 29 completed campaigns (pulled 9 August 2026; test records and still-running campaigns excluded). Pools ran from $3,500 to $203,262, entries from 11 to 3,899. The correlation between the two, measured on logs so the largest campaigns don’t dominate, is 0.12.

Sorted into four bands by pool size, median entries came out at 292, 359, 710 and 564, smallest pools to largest. There is a lift in there — the top half does better than the bottom half — but it’s a rough doubling of entries for something like a twenty-fold increase in prize money, and it isn’t even monotonic. The spread inside each band is far wider than the gap between bands. The biggest pool in the set, just over $203,000, returned 750 entries. A $50,000 pool returned 3,899; another at $49,900 returned 15.

Nor is it the number of winners rather than the size of the pot — we checked, and the correlation there is 0.21, no better.

Twenty-nine campaigns is small and none of it is a controlled experiment, so the honest read isn’t “prize money is irrelevant” — it’s that prize money is nowhere near the strongest thing in the equation, and something else is doing the heavy lifting. In our set it was reach and access. The campaigns at the bottom of the entry range were mostly trade activity, running to a few hundred venues or a wholesaler’s account base, where a couple of dozen entries is a reasonable result. The ones at the top ran through national retail with the offer visible where people were already shopping.

Which resolves the three campaigns at the top of this article. Their prize pools were identical. The number of people who could see and enter them was not.

What the prize budget actually has to achieve

Two things, and both are pass/fail rather than more-is-better.

The first is being worth the bother. A shopper decides quickly whether the reward justifies the effort of claiming it, and if it doesn’t, no amount of headline styling rescues it — that’s the insult threshold. A small-dollar reward sitting behind a receipt upload and a long form fails it however the offer is worded.

The second is being believable. A single major prize reads to most people as something that happens to someone else, which is why how the pool is divided is a separate decision from how big it is. Our set is too small to tell you the right split, and anyone who quotes you one with confidence is guessing — but it is a decision, and it usually gets made by whoever fills in the prize table last.

Once both are cleared, extra prize money isn’t fixing anything the shopper is weighing up. It’s just a bigger number sitting in the same place.

If the brief demands a big number, buy it rather than fund it

Sometimes a large headline prize genuinely is non-negotiable — the retailer wants it as the price of the feature, or the category is loud enough that a $15,000 draw disappears into it. When that’s the case, insure the prize rather than sit on the full liability. Prize indemnity insurance lets you advertise a prize that would be uneconomic to underwrite yourself, paying a premium against the odds of it being won instead of setting the whole amount aside. What you owe the winner doesn’t change; the ACCC expects prize and cash-back offers to run exactly as advertised. Only the cost of carrying it changes.

The money that comes back out of the pool has better places to be. Given what our numbers say about reach, on-pack real estate, shelf presence and retailer media are usually a better buy than the increment from a $20,000 major prize to a $30,000 one — as is every entry step you can delete.

So how much should a promotion prize be worth?

Enough to clear both thresholds for the job the promotion is actually doing, and not much more. A trial campaign wants breadth, because it needs a lot of people to act once. A basket-building campaign wants a reward that scales with spend. A data capture campaign needs less than most briefs assume, because an email address is a cheap thing to buy. Settling that before anyone names a prize does more for the budget than the argument about the major prize ever will.

It’s an awkward conversation to have with a brand team that has just had a bigger prize budget approved, and one Trevor Services ends up having fairly often. If there’s a prize pool sitting on your desk right now, we’re happy to have it with you.

One last number, because it’s the one that ends the argument fastest. Across those 29 campaigns, a thousand dollars of prize money bought a median of 22 entries. The best campaign in the set got 188 for the same thousand dollars. The worst got 0.18. Whatever explains a thousand-fold gap like that, it isn’t the size of the prize.

Eighty-nine of the promotions we’re tracking in Australian retail this morning are chance-based — 49 single prize draws, 40 instant wins. Two are cashbacks. That’s out of 137 live campaigns we log across FMCG, liquor, appliances and general retail; the remaining 46 are mostly gift-with-purchase, with a handful of money-back guarantees and collect-to-get mechanics.

It isn’t a reading of what shoppers want. It’s a reading of what a finance team will sign.

Our own cashback forecasts came in at roughly half

Here is the number that should change how you budget, out of the campaign records Trevor Services keeps. On an Electrolux kitchen bundle cashback, the pre-campaign claim estimate was around 1,700. Validated claims landed at 888. On a Westinghouse bundle running much the same structure, the estimate was roughly 2,600 and about 900 people claimed. Two independent campaigns, same category, both landing between a third and a half of forecast.

Be clear about what that does and doesn’t prove. We don’t publish sales denominators for client campaigns, so it isn’t a claim rate — two campaigns isn’t a law of nature either, and part of what it says is simply that the estimate was built optimistically. What it does say, reliably enough to budget on, is that a pre-campaign claim estimate is a ceiling and not a plan. If you’re building a promotional P&L on an appliance-style bundle, budget the full liability, model the likely spend well below it, and decide in advance what you’ll do with the difference.

Most brands book that difference as a saving. We’d argue it’s the most expensive line in the campaign.

An under-claimed cashback isn’t a saving

Some of the gap is people who bought without noticing the offer. Most of it, in our experience of watching claim funnels, is people who noticed, started, and stopped — the receipt photo was too dark, the model number was in the wrong place, the form asked for something they didn’t have to hand. That’s slippage, and slippage is genuinely what makes a cashback cheaper than a straight discount. But the money you didn’t pay out is money that did no work, and someone who abandoned a claim has learnt something about your promotions that shows up the next time you run one.

It’s what we call the Insult Threshold in The Shelf Truth, our promotions playbook: if the reward isn’t worth the effort of claiming it, you haven’t run a promotion, you’ve run a test of your customers’ patience. On a single $4 packet of biscuits, no cashback clears that bar — the claim takes longer than the money is worth, and no amount of form design fixes it. The way around it is aggregation: buy six, get $10 back. That works, but it’s a different promotion with a different job, and it only makes sense if the category is bought in multiples. On a $2,000 appliance bundle, $300 back is worth ten minutes and a photo of a receipt. The mechanic is good or bad relative to the price of the thing, and the appliance category runs on cashbacks for exactly that reason.

What’s the real difference between a cashback and a prize draw?

A cashback pays a fixed amount to every shopper who buys and submits a valid claim, so its total cost depends on how many people claim. A prize draw pays a large prize to a small number of entrants selected at random, so its total cost is fixed the moment the prize pool is set. One is a forecast; the other is a number you can put in a budget line and defend.

That difference, not shopper psychology, is what settles most mechanic debates in Australian planning meetings. Nobody has to justify a forecast that can’t move.

The compliance asymmetry runs the other way

The mechanic finance treats as the safe one is the one that carries a regulatory process.

A prize draw is a game of chance, which puts it in trade promotion lottery territory. In NSW, an authority is required where the total prize value exceeds $10,000, under the NSW Government’s trade promotion rules. In the ACT, a permit is required unless the total prize value stays at or below $3,000, per the ACT Gambling and Racing Commission.

This is less of a timing problem than people assume, and it’s worth knowing why. NSW issues an authority for one, three or five years covering multiple promotions, so a brand that promotes regularly pays the friction once. It’s the first-timer, or the brand whose authority lapsed in a restructure, who discovers the process three weeks out from a national on-pack. What the permit does bite on is change: once the promotion is running, the terms you lodged are the terms you’re stuck with.

A cashback has no element of chance. Everyone who qualifies gets paid, so there’s no lottery and no lottery permit; the obligations are consumer law ones about clear terms and honouring what you advertised. The mechanic with the unpredictable cost carries the lighter regulatory load, and the mechanic finance likes because its cost is fixed is the one with the paperwork. Our guide to Australian competition permits has the state-by-state detail.

If you’re running a draw, run it for a reason

Plenty of products can’t carry a cashback, and for those a chance mechanic is the honest answer. The job then is making the odds feel real rather than making the headline big. Our working rule — the Rule of Three, and it’s a heuristic from running these rather than a measured effect — is that one prize reads as impossible, three read as possible, and a hundred read as probable. A single $100,000 headline against a $6 product looks impressive on-pack and mostly rewards people who were buying anyway. We’ve argued that case at length in prize pool distribution models.

The other thing a draw won’t do for you is data. A prize draw gets you an email address and a stated intent. A receipt-validated cashback gets you the product, the retailer, the date and the price paid — verified purchase data you can plan the next campaign from. If your promotion has a data job attached to it, that difference is the whole decision.

And if the draw exists because you want a headline prize you can’t fund, that’s a financing problem with a financing answer: prize indemnity insurance lets you advertise a prize far larger than your budget for a premium you know up front. Trevor Services sets those up regularly, and it’s a better solution than shrinking the prize until nobody cares.

Pick the failure you can afford

A prize draw fails quietly. Entries come in low, the prize goes to someone who was buying anyway, and the campaign ends with nobody able to say much about what it did. The cost was capped and so was the upside, which is why it rarely gets a post-mortem.

A cashback fails in one of two directions. Everyone claims and you run past forecast — uncomfortable, but it means the offer worked. Or almost nobody claims, the finance report looks excellent, and you’ve quietly taught a slice of your buyers that your promotions aren’t worth their time. The second one costs more and is much harder to see, which is why it keeps happening.

So the position is this. On considered purchases the default prize draw is the wrong call, and it keeps winning the meeting because its cost is legible, not because it works better. If your product can carry a cashback, make the claim easy enough that people finish it, and read a low claim rate as a fault in the design rather than a windfall to bank. Trevor Services builds both kinds every week, and we’re happy to talk through which failure you’re buying.

The prize draw that’s genuinely right for a campaign survives that conversation easily. It’s the one nobody can explain, beyond the fact that the number was easy to sign off, that costs you a quarter.

Somewhere in most cashback planning meetings, the offer gets costed at face value. A $100 cashback on a $1,000 appliance goes into the spreadsheet as $100 a unit, the same as a discount would. Anyone who has processed the claims knows that’s not how it plays out — a meaningful share of the people who buy on the promise of a cashback never get around to claiming it. The gap between the two numbers has a name, decades of research behind it, and more influence over a cashback budget than any other single figure. It’s also the number most likely to be missing from the plan.

What is slippage in a cashback promotion?

Slippage is the percentage of eligible buyers who never claim the cashback they’re entitled to. Because unclaimed cashbacks cost the brand nothing, a cashback’s true cost is its face value multiplied by the redemption rate — not the face value itself — which is why a cashback almost always costs less than a discount of the same headline amount.

The industry sometimes calls the same thing “breakage”, borrowing the term from gift cards. Either way, it isn’t a loophole and it isn’t something to engineer. The receipt goes in the bin, the claim window closes, the form gets abandoned halfway — none of it because you designed it that way. Your job is to forecast it accurately, not farm it. And forecasting it is harder than most planning meetings assume, which is the part of this that actually deserves your attention.

The arithmetic: what a cashback actually costs

A discount reaches 100% of buyers, every time, whether they noticed the promotion or not. A cashback reaches only the buyers who claim — and claim rates sit well below what most planners assume. The best public data is a 2023 study by Tremendous and The Decision Lab, which found digital rebates of $20 or less were claimed only around 38% of the time, rising to roughly 50% for $50 rebates. Run that through the spreadsheet: at the study’s observed rate, a $50 cashback costs about $25 a unit in redemptions, where a $50 discount costs the full $50 on every sale. Same headline offer to the shopper, half the redemption cost to the brand.

The academic literature explains why the gap is so persistent. Scott Gilpatric’s Marketing Science paper on slippage in rebate programs ties it to present-biased preferences: the purchase happens now, the claiming effort comes later, and later is where good intentions go to die. The same study data shows claim rates climbing as the money gets bigger — which is why the cashbacks running in the Australian market right now cluster in high-ticket categories: Sony offering up to $1,000 on selected cameras and lenses, LG up to $300 on TVs through Betta, OM System (the old Olympus camera business) up to $500. Nobody runs a $5 cashback on purpose.

How do you forecast a redemption rate?

A redemption rate is a forecast, not a constant, and it moves with five things: the value of the offer relative to the purchase, the friction in the claim process, the length of the claim window, how quickly the money arrives, and who the buyer is. A $15 cashback claimed through a clunky form with a 30-day window and a six-week EFT payout will slip enormously. A $300 appliance cashback claimed by scanning a QR code, uploading a receipt and receiving a PayID payout inside a day will not. Every choice you make about the claim journey moves the rate — which means slippage is partly a design outcome, and you should know which way your design is pushing it before you commit a budget number.

Here’s how wrong the forecasts get, from our own claim queues at Trevor Services. Two recent appliance bundle cashbacks we processed: one was budgeted for roughly 1,700 claims and closed under 900 — 52% of forecast. The other was budgeted for about 2,600 and also closed under 900 — 35% of forecast. Both promotions came in far cheaper than planned, which sounds like good news until you notice that the same forecasting error in the other direction would have blown the accrual by two to three times. If operators with campaign history on hand can miss by that margin, a redemption rate pulled from instinct in a planning meeting isn’t a forecast, it’s a guess with a spreadsheet cell.

Which is why this becomes a finance conversation, not just a marketing one. The redemption forecast sets the liability you accrue, and a promotion that out-redeems its forecast doesn’t fail loudly — it fails in the accruals, months later, when finance asks why claims are still coming in. The honest approach is to budget at a conservative redemption rate, track actual claims weekly against the forecast, and re-accrue as the pattern emerges; the first fortnight of claims data usually tells you where the campaign is heading. For brands that can’t carry the tail risk of over-redemption, sales promotion insurance exists precisely to cap it: you pay a fixed premium and the insurer wears the variance. The brands that get burned by slippage aren’t the ones using it — they’re the ones who never put a researched number on it. And the best predictor of your next claim rate isn’t the planning meeting’s instinct; it’s what similar offers actually did, which is exactly the history worth consulting before you commit the budget line.

The line you can’t cross: slippage and the ACCC

There’s a version of this thinking that tips into misconduct, and it’s worth being blunt about where the line sits. Budgeting for the fact that some people won’t claim is legitimate. Designing the claim process so that people can’t claim — burying conditions, shrinking windows, adding gratuitous steps — is not, and it’s squarely in the regulator’s sights. The ACCC’s guidance on cash back offers is plain: conditions and limitations must be clear to the consumer before purchase, and a business that offers a rebate must intend to honour it as offered. The regulator has been warning brands about undisclosed cashback conditions for years, and a promotion that quietly relies on entrapment rather than forgetfulness is a complaint waiting to be lodged.

The practical test is the one we call the Insult Threshold in The Shelf Truth, our promotional strategy guide: if the effort of claiming isn’t worth the reward, you haven’t saved money, you’ve insulted a customer who did exactly what your advertising asked. High slippage driven by a low-value offer or a hostile claim process isn’t a budget win — it’s a signal the promotion shouldn’t have run in that shape at all. The cashbacks that work are the ones where claiming is easy, payment is fast, and the slippage that remains is the genuine, unforced kind.

Forecast it, don’t farm it

Slippage is the reason a cashback can deliver a $100-off message for materially less than $100 a unit. It’s also the least reliable number in the plan — our own claim queues show forecasts missing by half — which means it deserves the most scrutiny, not the least. Put a researched number on it before launch, design the claim journey deliberately rather than accidentally, track actuals weekly, and stay on the right side of the ACCC’s line. If you’re building a cashback budget and want to pressure-test the redemption assumptions against real campaign history rather than instinct, we’re happy to talk it through. The slippage will take care of itself — it always does. The forecast won’t.