Two loyalty cards, the same eight visits needed for a free round — and yet one card fills up almost twice as often as the other. The only difference: two stamps that were already on it.
In 2006, an American car wash tested two loyalty cards on hundreds of customers. One card had eight boxes and had to be stamped from zero. The other had ten boxes, two of which were already stamped when it was handed out — so it also had eight washes left to go. Same effort, same reward, a completely different result: the card with the head start was filled to completion almost twice as often.
That's the endowed progress effect (Nunes & Drèze, 2006), one of the best-documented findings in the psychology of loyalty programs. Its sibling, the goal-gradient effect (Kivetz, Urminsky & Zheng, 2006), explains why visits come faster and faster as a guest gets closer to the reward — a pattern first described a century earlier in animals navigating a maze, and one that transfers flawlessly to people holding a loyalty card.
Both effects play out on every loyalty card a hospitality business hands over, whether the owner has thought about it or not. This guide walks through the seven numbers that decide whether your card feels like a sprint or a marathon to a guest: how many stamps you ask for, whether you give a few away for free, and where businesses undo all of it themselves by letting a card expire or forcing a guest to re-register.
At the bottom, plug your own card into the calculator: the number of stamps, the head start you're considering, how many cards you hand out per month, and what a visit is worth to you. You'll get the expected rise in completion, what that means in extra visits and revenue, and what it costs you in extra free rewards. Everything runs in your own browser: nothing is sent anywhere and nothing is stored.
Why a card that starts at zero is a missed opportunity, not a mistake
A loyalty card is, at its core, a piece of arithmetic: stamps needed minus stamps earned. For a guest, it isn't arithmetic — it's a feeling of distance. Two stamps already on the card when it's handed over don't feel like "two free washes" — they feel like "I'm already on my way". That feeling exists independently of the eight visits still actually required, and it's exactly what the endowed progress effect measures.
The reason lies in how people value progress toward a goal: not linearly, but relative to what's left to go. The first stamps on a card feel the heaviest — the card is still nearly empty, the reward far away, and many guests drop off right here. A two-stamp head start moves the starting point past that heaviest stretch, without making the card any cheaper: the number of visits you still have to pay for stays exactly the same.
That's exactly why the endowed progress effect isn't the same as giving a discount. A discount lowers the price; a head start moves the starting point of the count. Both cost you something, but only the second changes how a card feels from the very first visit — and that feeling is precisely what decides whether a guest keeps the card in their wallet or forgets it after one stamp.
The Ultimate Guide Guest Experience & Concept: 6 Steps to Unforgettable Evenings From first impression to last bite: everything guest experience can do for your revenue, in one guide. Open the guideThe 7 numbers behind a card that fills itself
They're presented in the order you'd use them to design your own card: first the evidence, then the mechanism, then its limits.
1. Two identical cards, two completely different results
Nunes and Drèze's study ran at an American car wash, with more than three hundred customers. Group A received a card with eight boxes: eight washes for a free ninth. Group B received a card with ten boxes, two of which were already stamped — that group also still needed eight more paid washes before the free one. The effort was identical for both groups.
The result wasn't. Card A was fully filled by 19% of customers. Card B, with the two free stamps on it, by 34%. Almost a doubling, for a head start that cost the car wash nothing beyond two stamps on a card it was handing out anyway.
Line the cards up side by side and the reason is obvious: card A looks like a blank page at the first visit. Card B looks like a card that's already 20% of the way there. That difference in feeling — not in arithmetic — is the entire effect.
Card A starts empty. Card B gets two stamps for free at handover — and then asks for exactly the same eight visits. The only difference is what the guest already sees on it.
Both cards require exactly 8 paid visits
Source: Nunes & Drèze (2006), Journal of Consumer Research — a controlled test at an American car wash, more than three hundred customers. The head start changes nothing about the arithmetic, only about how the first visit feels.
2. Why your brain feels the distance, not the number
Ask a guest how many visits are left on either card and most will work it out correctly: eight, on both. Yet people don't behave according to that number — they behave according to how far the card looks from finished. That's not a miscalculation; it's how people judge goals: relative to what's already been achieved, not absolute to what's still needed.
The same mechanism explains why the first few stamps on any card weigh the heaviest. At zero stamps, a hundred percent of the way is still to go; after the first few stamps — free or earned — that percentage has already dropped, and every following stamp feels lighter than the last. A two-stamp head start simply skips that heaviest stretch.
For a restaurant, this means the question isn't how much discount do I give away, but where do I let the guest start feeling the card is already theirs. That's a design question, not a pricing question — and it's exactly why a loyalty card with a head start doesn't have to cost more than one starting at zero.
3. The sprint to the free one
Kivetz, Urminsky and Zheng tested something additional in 2006, at a coffee shop: how quickly guests came back as their card filled up. On an ordinary ten-stamp card, the visit pace measurably accelerated in the last few stamps before the free coffee — guests who used to return every so many days came back more often as the reward approached.
Among guests with a head-start card — twelve boxes, two already stamped, also ten net visits to go — that same acceleration was there too, and stronger. The head start made the felt endpoint closer, and that sense of proximity is exactly what drives up the pace. It's the same pattern described much earlier in animals that ran faster the closer they got to food in a maze — the goal-gradient effect, here convincingly demonstrated for the first time on an everyday loyalty card.
For a loyalty card, this means two things at once: a guest with a nearly full card doesn't just come back more reliably, but also faster. That's the moment a well-timed reminder — a card pulled out at checkout, a message at the last two stamps — works hardest, because the guest is already the most receptive to it.
The closer a guest gets to the reward, the faster the next stamp follows — on an ordinary card, and even more so on a card with a head start.
An illustration of the pattern Kivetz, Urminsky & Zheng (2006) measured at a coffee shop: visit pace steadily accelerates as the card fills up, and that curve is even more pronounced for guests who started with a head start. The curve shows the direction of the effect, not exact figures for your business — how strongly this happens for you depends on your own visit rhythm.
4. How much of a head start is enough
The car wash gave away two of ten stamps: 20% of the card. That's the only ratio the study actually tested, and it's exactly why this guide doesn't pretend a 60% head start delivers three times the uplift of a 20% one. That relationship has never been measured.
What does follow logically from the mechanism: a one-in-ten head start is probably too small to feel like a real leap forward. An eight-in-ten head start leaves only two real visits — at that point the card isn't really a loyalty card so much as a gift voucher with a detour, and the guest has little reason to come back for the feeling of progress itself.
The calculator at the bottom of this guide therefore sticks to what's been measured: the gain scales up to a head start of 20% of the card, and is never increased beyond that. Give away more, and you're doing it on faith, not evidence — and you do pay for every free box, in rewards redeemed sooner.
5. Where restaurants undo this themselves
The easiest mistake is doing nothing: a card that starts at zero gains nothing from this effect. That's not a disaster — it's simply the card from group A at the car wash, with its 19%. But it is a missed opportunity that two stamps at handover could have closed.
The second mistake is bigger: letting a card expire. Loss aversion works against you here instead of for you — a guest who loses four of eight stamps to an expiry date doesn't feel "four stamps gone", they feel "I'm losing something that was already mine". That feeling is stronger than the endowed progress effect ever was, and it works in the opposite direction: toward irritation, not toward another visit.
The third mistake is technical: a new point-of-sale system that forces guests to re-register and resets their balance to zero in the process. For the business, that's a migration. For the guest, it's exactly the same loss aversion as an expired card, only this time caused by the owner. Every switch to a new system deserves one line of policy: existing stamps carry over, always.
6. How many stamps is too many
Both studies used cards of eight to ten boxes, and that's no coincidence: it's the length at which a loyalty card still feels achievable within the time a guest actually thinks about it. A card that only pays out after twenty visits asks a guest who might come by once a month for nearly two years of patience — and most of those cards end up in a drawer before they're full.
That's a different mechanism from the endowed progress effect, but it works in the same direction: the closer the reward feels from the start, the greater the chance a card gets finished. A shorter card with a small head start wins on both fronts at once.
Rule of thumb: fit the card to how often a guest realistically returns. For a daily lunch spot, ten visits can happen within a few weeks; for a restaurant a family visits once a month, eight visits is already a year and a half. Shorten the card, or give a bigger head start — both shrink the road to feeling nearly done.
7. The reward matters just as much as the card
The endowed progress effect is about the card's starting point, not about what's waiting at the end. Yet that reward decides whether a guest thinks finishing is worth the effort: a free coffee after ten paid coffees is a different story from a free main course after ten full meals, even if the card is ten boxes long in both cases.
A reward that feels too small relative to the effort undermines the effect before it can even work: a two-stamp head start doesn't help if the remaining eight visits for a reward that barely amounts to anything simply don't feel worth it. Always weigh the card length and the reward together, never separately.
And the card doesn't have to be the same every time: a printed loyalty card with a fixed reward works well for a recurring dish or drink, a gift voucher works better when the reward's value needs to scale with what a guest normally spends. Both are the same principle in a different coat.
Run the numbers for your own card
Fill in how many stamps your card asks for, how many you give away for free at handover, and how many cards you hand out per month. You'll get the expected rise in completion — capped to what the study actually measured — and what that means in visits, revenue and extra rewards.
The pre-filled values describe an average neighbourhood restaurant with a simple loyalty card. Overwrite them with your own.
Head-start calculator
Stamps, head start, cards per month — and what it earns you.
—
The 20% cap and the 1.79× multiplier come directly from Nunes & Drèze (2006). This model scales linearly up to that cap and never goes beyond it — bigger head starts aren't modelled as extra gain, even though they do cost extra free stamps. Everything runs in your browser: nothing is sent or stored.
Two things to know when reading this. The extra filled cards cost you one extra reward each — those guests, without the head start, would probably never have finished their card, so that cost mostly didn't exist before. Against that, every completed card also represents the paid visits that led up to it: exactly the number of stamps minus your head start.
The net figure at the bottom weighs those two against each other. At most realistic head starts (one to two stamps on a card of eight to ten), that figure is comfortably positive, because the revenue from the extra visits far outweighs the cost of one free reward.
What to do this week, this month and this quarter
Redesigning a loyalty card doesn't have to be a project. This order delivers the most for the least effort.
This week — add the head start
- Count how many boxes your current card has, and compare that with how often an average guest realistically comes back.
- Pre-stamp one or two boxes on every new card you hand out from now on — change nothing else.
- Run your head start through the calculator above and note what you expect in extra filled cards.
- Don't change anything about cards already in circulation: the head start is for new cards, not retroactive.
This month — measure what changes
- Track separately how many cards with and without a head start are redeemed for their reward.
- Ask the floor whether guests notice anything about the new card, without leading the question.
- Check whether the reward still feels worth it relative to the number of visits remaining.
- Adjust the head start if it turns out to be one stamp too few, or too many: you're looking for the point where the card feels like a leap, not a gift.
This quarter — protect what you've built
- Drop an expiry date on your loyalty card, or extend it substantially: expired stamps cost you more in goodwill than they save.
- For every switch to a new point-of-sale or loyalty system, agree upfront that existing stamps carry over.
- Put the calculator next to your customer lifetime value calculation every quarter — a loyalty card that fills up more often raises exactly that number.
- Consider a gift voucher alongside the loyalty card for bigger spends: the same head-start principle works there too.
The card costs you nothing extra — the starting point does
The endowed progress effect is one of the rare pieces of guest psychology that cost literally nothing to apply: two stamps on a card you're handing out anyway change nothing about what a guest ultimately has to do to earn their reward. The only thing that changes is how far away that reward feels from the first visit.
Get it wrong and you lose the same amount in goodwill: a card that expires, or a system switch that wipes stamps, triggers loss aversion instead of progress — and that feeling is stronger than any head start you've ever given.
Start small: two stamps on a card of eight to ten, no expiry date, and a reward that feels worth it. After a quarter, work out how many more cards get filled, and weigh that against what it cost you in extra rewards. For most businesses, that's the easiest positive number this guide produces.