Not legal advice: This article covers marketing practice and general legal awareness of Israeli consumer-protection rules around customer clubs and benefit programs, not a legal audit. If you run a paid or points-based customer club, confirm your current notice and voucher-validity obligations with a qualified lawyer.
A punch card is one of the oldest tricks in marketing — buy nine, the tenth is free — and it's easy to assume the psychology behind it is as simple as the card itself. It isn't. Two of the best-known findings in the field involve giving two groups of real customers the exact same reward for the exact same number of purchases, and watching completion rates move anyway, just from how the card looked on day one. And a separate, less-quoted branch of the same research asks a blunter question: once you strip away the psychology of punch cards and status tiers, does a loyalty program actually make people buy more — or does it mostly hand free stuff to customers who were going to keep buying anyway?
We went and checked the originals: the 2006 field experiment that handed out 300 real car-wash cards to test whether an "artificial" head start changes behavior, a separate 2006 study that tracked 949 real, redeemed café loyalty cards and nearly 10,000 actual coffee purchases, a 2009 study on what tiered "VIP" status structures do to how special people feel, the 1997 paper that first argued loyalty programs are oversold, a 2009 French study of 451,000 real purchase records that tested the same question with panel data, and — because this matters for any Israeli business running one — the 2017 rule that controls what a customer club has to tell members before it changes or ends their benefits.
Key takeaways
- In a real field experiment, 300 car-wash customers got loyalty cards requiring 8 paid washes for a free one. Customers given a 10-stamp card with 2 stamps already filled in completed it 34% of the time, versus 19% for an empty 8-stamp card requiring the identical 8 purchases — nearly double, from an artificial head start alone.
- In a real café loyalty program tracked by researchers (949 completed, redeemed cards recording close to 10,000 actual coffee purchases), members bought their next coffee an average of 20% faster as they got closer to their free one, and completed their cards about 16% faster — roughly 5 days sooner — than their own early-program pace predicted.
- A 2009 study found tier structure itself shapes status, independent of what each tier actually offers: adding a lower tier below an existing one makes the tier above it feel more elite, while growing the size of the top tier dilutes everyone's status in it — and "gold" or "silver" labels alone signal a more selective program even with no other information.
- Not every finding is flattering. A 1997 Sloan Management Review paper argued loyalty programs rarely grow the pool of loyal customers, since perceived brand value — not program membership — is what drives extra spending; a 2009 study of 451,000 real purchase records from 2,150 French shoppers over 156 weeks found heavy buyers mostly self-select into a program rather than being created by it, with little measurable change in their actual purchase behavior.
- Since December 2017, Israeli law (Section 16א of the Consumer Protection Law) has required any business running a customer club or benefits program — paid or free — to give members 3–4 months' notice before changing or ending their benefits, and to let them redeem what they've earned until the change takes effect; separately, a gift or credit voucher must stay valid for at least 2 years and a store purchase voucher for at least 5.
Does giving customers an artificial head start actually make them finish a loyalty card?
The cleanest test of this comes from Joseph Nunes and Xavier Drèze, "The Endowed Progress Effect: How Artificial Advancement Increases Effort," published in the Journal of Consumer Research in March 2006 — a real field experiment, not a lab survey. In April 2004, the authors randomly handed out 300 loyalty cards at a professional car wash. Every card required exactly 8 paid washes to earn a free one. Half the cards had 8 empty spaces. The other half had 10 spaces, with 2 already stamped — meaning both groups needed to buy exactly 8 more washes for the same free wash. The only difference was how far along the card looked on the day a customer got it.
That single difference nearly doubled completion. Customers given the 10-space card with a 2-stamp head start finished it 34% of the time; customers given the plain 8-space card finished it only 19% of the time, despite needing to make the identical number of purchases. The authors also found the head-start group didn't just finish more often — they finished sooner, buying their washes at a faster pace than the group starting from zero.
The authors call this the "endowed progress effect": people respond to how close a goal looks, not only to how much work objectively remains. It's a genuine behavioral finding from a real business, with randomly assigned customers and an identical reward on both sides — not a framing trick that only works in a lab.
Do people really speed up their purchases as they get closer to a reward?
A separate, equally rigorous field study tested a related but distinct question: not whether a head start helps, but whether people genuinely accelerate as they approach a goal they're already working toward. Ran Kivetz, Oleg Urminsky and Yuhuang Zheng published "The Goal-Gradient Hypothesis Resurrected: Purchase Acceleration, Illusionary Goal Progress, and Customer Retention" in the Journal of Marketing Research in February 2006. The data came from a real café loyalty program at a university campus café: members needed 10 stamps for a free coffee, and the researchers tracked actual purchases, not a survey about intentions.
The scale is what makes this one hold up: 949 completed, redeemed ten-stamp cards, recording close to 10,000 real coffee purchases, modeled as 29,076 individual purchase-day observations in a hazard-rate model built specifically to detect whether buying sped up as members got closer to their free coffee. It did. The average time between purchases shrank by about 0.7 days — roughly 20% — from a member's first stamp to their last (t = 2.6, p < .05), and the statistical model confirmed the effect wasn't explained away by time trends, habituation, or who happened to buy more coffee overall.
20%
the average acceleration in how fast real café-loyalty members bought their next coffee, from their first stamp to their last — measured across 949 completed, redeemed cards and close to 10,000 actual purchases.
Source: Ran Kivetz, Oleg Urminsky & Yuhuang Zheng, "The Goal-Gradient Hypothesis Resurrected: Purchase Acceleration, Illusionary Goal Progress, and Customer Retention," Journal of Marketing Research, Vol. 43 (February 2006), pp. 39–58
The same acceleration showed up as a completion-time effect too: the average card was finished in 24.6 days, versus the 29.4 days it would have taken at a member's own early-program pace — about 16% faster, or nearly 5 days sooner. And in a direct echo of the car-wash result above, the same paper separately found that members given a 12-stamp card with 2 "bonus" stamps already attached completed their 10 required purchases faster than members given a plain 10-stamp card — the illusionary-progress effect replicating, independently, in a second real program.
Do tiered "VIP" status levels actually make top customers feel special?
Xavier Drèze and Joseph Nunes — the same researchers behind the car-wash study — turned to a different question in "Feeling Superior: The Impact of Loyalty Program Structure on Consumers' Perceptions of Status," published in the Journal of Consumer Research in April 2009: once a program has tiers (think "Silver / Gold / Platinum"), what actually controls how special the top tier feels? Their answer is about structure, not just rewards.
Two findings stand out. First, adding a subordinate (lower) tier beneath an existing structure makes the tier immediately above it feel more elite, even when nothing about that tier's own requirements or benefits changes — status is relative to what's visibly below you, not just to the general population. Second, the effect runs the other way at the top: increasing how many members qualify for the top tier dilutes the status everyone in it feels, even though each individual member still did the same thing to get there. The authors also found that status-laden labels — "gold," "silver" — signal a more selective program on their own, independent of any other detail, and that customers who don't personally qualify for status still prefer programs structured with multiple tiers over a single flat one.
The practical read: a flat "member / non-member" structure leaves status entirely to how the program is marketed, while a cheap, low-bar entry tier added beneath an existing top tier can raise how elite the top tier feels without changing a single one of its actual requirements or rewards — and widening who qualifies for the top tier is a real tradeoff against how special it still feels to the people already in it, not a free way to make more customers happy at once.
Does a loyalty program actually grow revenue — or mostly reward people who were buying anyway?
Not every well-regarded study is flattering to loyalty programs, and the brand's own standard here is to report what doesn't hold up along with what does. The earliest sharp pushback is Gary Dowling and Mark Uncles, "Do Customer Loyalty Programs Really Work?", published in Sloan Management Review in 1997. Their argument: loyalty programs are unlikely to meaningfully grow the share of genuinely loyal customers a brand has, because it's the brand's perceived value — not program membership on its own — that drives a customer to spend more. Rewarding customers who'd have kept buying regardless doesn't create new demand; it mostly redistributes margin to people who were already there.
A later, data-heavy test of the same question is Lars Meyer-Waarden and Christophe Benavent, "Grocery Retail Loyalty Program Effects: Self-Selection or Purchase Behavior Change?", published in the Journal of the Academy of Marketing Science in 2009 — 451,000 real purchase acts from 2,150 consumers in Angers, France, tracked over a 156-week BehaviorScan panel. Their finding: customers who already bought more, and lived closer to the store, were the ones who joined the loyalty program earliest — a self-selection pattern, not a causal one. Once enrolled, heavy buyers' actual purchase behavior barely moved, with only a slight loyalty increase showing up 6–9 months in, and the program failed to pull in the most valuable prospects of all: heavy buyers of competing stores.
A third study complicates a purely negative read. Yuping Liu, "The Long-Term Impact of Loyalty Programs on Consumer Purchase Behavior and Loyalty," Journal of Marketing, 2007, used longitudinal data from a U.S. convenience-store franchise's real loyalty program. Heavy buyers behaved exactly as Meyer-Waarden and Benavent later found — quick to redeem rewards, unmoved in their underlying behavior. But customers who started out as light or moderate buyers told a different story: they gradually bought more and grew more loyal over time, with the clearest shift inside the first 3 months of joining and continued, slower growth over roughly two years — plus a secondary effect where the program nudged them into other parts of the business, lifting cross-sell revenue along the way.
| Study | What it actually found |
|---|---|
| Dowling & Uncles, 1997 — Sloan Management Review (conceptual argument, not new data) | Loyalty programs are unlikely to meaningfully grow the pool of loyal customers; perceived brand value, not membership, drives extra spending |
| Meyer-Waarden & Benavent, 2009 — Journal of the Academy of Marketing Science (451,000 purchase acts, 2,150 shoppers, 156 weeks) | Heavy buyers self-select into the program early rather than being created by it; little change in their actual behavior; program fails to attract competitors' heavy buyers |
| Liu, 2007 — Journal of Marketing (longitudinal convenience-store panel) | Heavy buyers' behavior stays flat, but light/moderate buyers gradually buy more and grow more loyal — clearest shift within 3 months, continuing over ~2 years, plus cross-sell gains |
Read together, the honest synthesis isn't "loyalty programs don't work" — it's that they mostly don't work the way the pitch implies. They're weak at converting your already-heaviest buyers into spending more, and weak at pulling in a competitor's best customers. Where the evidence is actually encouraging is on your light-to-moderate buyers — the segment a program seems most plausibly able to move, and the one a simple "first purchase in 90 days" filter in your own customer data can isolate before you spend a cent on rewards.
What does Israeli consumer-protection law require from a customer club?
Since 7 December 2017, Amendment 53 to the Consumer Protection Law, 5741-1981 (adding Section 16א) has regulated what it calls a "benefits program" — defined broadly as any ongoing relationship where a consumer hands over their details to join a club-type program run by a business, in exchange for benefits or accruing rights. That definition covers a standard points-based customer club whether or not joining costs money — the rule applies to free benefit programs too, not only paid memberships.
The core obligation is advance notice. A business that wants to change or end a benefits program has to notify members 3–4 months before the change takes effect, and has to let them keep redeeming what they've already earned from the moment that notice goes out until the change (or the program's end) actually happens. For a program advertised with a fixed term, the business can't change or end it early, and still owes the same 3–4 months' notice of the member's right to redeem before the fixed term itself runs out.
Separately, if a club's points convert into a voucher rather than a direct discount, Sections 14ז–14ח of the same law set minimum validity periods: a gift voucher or a credit voucher (such as one issued for a cancelled purchase) must stay valid for at least 2 years from the date it's issued, and a general store purchase voucher (a תו קנייה, closer to prepaid store credit) for at least 5 years. Neither can require the original receipt to redeem, and neither can be written in a way designed to fade or become illegible.
Worth being precise about a separate, unrelated-sounding law that also applies here: because a customer club keeps a database of members' personal details, it's also a "database" under the Privacy Protection Law, 5741-1981 — and Amendment 13 to that law (not the Consumer Protection Law — a genuinely common mix-up, since both are called "Amendment 13" to their own, different statutes) took effect on 14 August 2025 with its own separate notice, access, correction and security obligations. We cover that law's requirements in full in our Privacy Law Amendment 13 article rather than repeating it here — a running customer club needs to satisfy both laws at once, not just the consumer-protection notice rule above.
How do you build a points-based loyalty flow that pays out on real purchases and doesn't quietly break the notice rule?
- 1Track the balance on the contact itself with a number custom field (e.g. loyalty_points), not in a spreadsheet — so every automation, segment and message can read and update it live.
- 2Update the balance off the Sale recorded trigger, not a manual click: the trigger fires the moment a purchase is logged to your Sales log, so points accrue on real revenue the same way the research above measured it, not on a cart someone abandoned.
- 3Use a Condition node to check the running balance against each tier's threshold, and update the contact's tags (e.g. tier: silver, tier: gold) the moment a threshold is crossed — tags are what your segments, campaigns and the channel router filter on everywhere else in oToK.
- 4Keep the tier ladder short and the step to the next one visible in the message itself ("120 of 150 to Gold") — the goal-gradient and endowed-progress findings above are both about visible proximity to a reward, not the reward's size.
- 5If a reward tier pays out as a voucher rather than a running discount, generate it as a unique, single-use coupon code per contact — the same mechanics covered in our coupon-code-marketing article — and keep its validity at or above the legal minimums covered above rather than setting an arbitrary shorter expiry.
- 6If you ever need to shrink or end the program, build the notice as its own flow first: a Manual run that tags everyone currently enrolled and sends the change by email with automatic WhatsApp fallback the same day, then a Wait node that holds the actual tier or reward logic update for the legally required 3–4 months before it's allowed to fire.
None of this needs a dedicated "loyalty" module — it's the same trigger, Condition, tag and coupon-code building blocks used elsewhere on the platform, chained so the points balance, the tier tags and the reward payout all stay driven by the Sale recorded trigger instead of a number someone has to remember to update by hand.
Frequently asked questions
- Does giving customers a head start on a loyalty card actually work, or is that a marketing myth?
- It's real and it's dramatic: in a 2006 field experiment, 300 real car-wash customers all needed exactly 8 more purchases for a free wash, but those given a 10-stamp card with 2 stamps already filled in completed it 34% of the time, versus 19% for a plain 8-stamp card — despite needing the identical number of purchases either way. Researchers call it the endowed progress effect.
- Do people really buy faster as they get closer to a loyalty reward?
- Yes, measured on real purchases, not survey answers: a 2006 study tracked 949 completed, redeemed café loyalty cards recording close to 10,000 actual coffee purchases and found members bought their next coffee about 20% faster, on average, as they approached their free one, and finished their cards roughly 16% faster than their own early-program pace predicted.
- Do tiered VIP levels actually make top customers feel more special?
- Structure matters as much as the reward itself, per a 2009 study: adding a lower tier beneath an existing one makes the tier above it feel more elite even with no change to its own requirements, while letting more people into the top tier dilutes how special it feels to everyone already there — and labels like "gold" or "silver" signal selectivity on their own.
- Does a loyalty program actually increase sales, or mostly reward people who'd have bought anyway?
- The evidence is mixed, and we're reporting the skeptical studies rather than only the flattering ones: a 1997 paper and a 2009 study of 451,000 real purchase records both found heavy buyers mostly self-select into a program rather than being created by it, with little change in their actual behavior. A separate 2007 study of a convenience-store chain found the same thing for heavy buyers — but found light and moderate buyers genuinely did buy more and grow more loyal over time, with the clearest shift inside 3 months of joining.
- What does Israeli law require if I run a customer club or points program?
- Since December 2017, Section 16א of the Consumer Protection Law requires 3–4 months' advance notice before changing or ending a benefits program — paid or free — with members able to keep redeeming earned benefits until the change takes effect. If points convert into a voucher, it must stay valid at least 2 years (a general store purchase voucher, at least 5). A customer club's member database separately falls under the Privacy Protection Law's own Amendment 13, in effect since August 2025 — a different law from the Consumer Protection Law despite the shared amendment number.
