Do Loyalty Rewards Actually Work? What a Stamp Card Really Costs You

For the trader with a stall, a shop, or a table in the market. The card that keeps getting stamped is not always the deal it looks like.

The question every shop owner asks sooner or later

A rep comes round, or a neighbour tries it first, and soon you are wondering the same thing: should you start a loyalty card? Buy nine, get the tenth free. A points app. A small cashback. It feels like the modern way to keep customers, and the cards do get stamped, so it looks like it is working.

Here is the harder question underneath. Cards getting stamped tells you people are buying. It does not tell you the scheme is making you money. Those are two different things, and the gap between them is where a lot of small profit quietly disappears.

Quick answer

Key takeaways

• Loyalty is two things, and rewards only touch one. There is repeat buying (how often people buy) and genuine preference (whether they actually favour you). Cards and points nudge the first; they do little for the second³.
• A “buy-9-get-1-free” card is about a 10% discount in disguise, and the hit to your profit is far bigger than 10%.
• A reward given for buying you already had is money donated. The whole case for a scheme rests on the extra buying it truly causes.
• Rewards amplify loyalty; they cannot manufacture it. Bolt a scheme onto a shop people already like and it helps. Bolt it onto a shop they merely tolerate and it just subsidises them until a sharper offer opens across the road.

---

SEE — the card that keeps getting stamped while your margin shrinks

One-line takeaway: a busy card is a measure of buying, not of profit, and the two can move in opposite directions.

Picture a kiosk that sells fast-moving goods in a busy Nairobi estate. The owner starts a stamp card: buy nine, get the tenth free. Regulars love it. The cards fill up, the tenth item goes out free with a smile, and it feels like loyalty in action.

Three months on, something is off. Turnover looks the same as before. The same familiar faces still come, but the month-end profit is thinner, and the owner cannot quite say why. Then a bigger shop opens down the road with its own sharper promo, and half those “loyal” card-holders drift over without a second thought.

That is the pattern worth seeing before you print a single card, and it is worth giving it a name so you can catch it next time: Rewards Buy Purchases, Not Loyalty. The customers who filled those cards were mostly people who were already buying. The scheme did not win them. It just handed them a discount for doing what they were going to do anyway, and it gave away real profit to do it. Once you have named it, you will recognise it everywhere: whenever a reward is being redeemed by people who never needed convincing, you are running this pattern.

You may be running this pattern if:

• The cards keep getting stamped, but your profit per month is flat or falling.
• The same regulars redeem rewards, month after month, that they always came for.
• A rival’s promo pulls your “loyal” customers away almost overnight.

---

UNDERSTAND — why rewards buy the shallow kind of loyalty

One-line takeaway: repeat buying is not the same as real preference, and a free-item deal is a price cut wearing a friendly costume.

Marketing researchers split loyalty into two parts. One is repeat patronage: how often a customer actually buys from you. The other is relative attitude: how much they genuinely prefer you over the shop next door. When someone buys from you often and truly prefers you, that is real loyalty, and it holds. When someone buys often but has no real preference, buying only from habit or because a card locks them in, that is what researchers call spurious (false or fragile) loyalty³. It looks the same on a busy day. It behaves very differently the moment a better deal appears.

A stamp card is built to lift repeat buying. It does almost nothing for genuine preference. So the loyalty it manufactures is mostly the fragile kind. This is not a fringe view. Reviewing the evidence, researchers found loyalty schemes tend to increase your costs without creating much extra loyalty, because so much of the reward flows to customers who would have bought anyway¹. Later work reached the same conclusion: in everyday repeat-purchase markets, buying is driven more by habit and convenience than by strong preference, so a programme’s power to lift demand is more limited than owners hope². When one real programme was measured carefully against normal buying patterns, the extra loyalty it produced was weak⁴.

There is a second thing to understand, and it is pure arithmetic. A free-item scheme is a discount in disguise, and the discount lands straight on your margin, not on your price.

What a buy-9-get-1 card really costs

Take that kiosk. An item sells for KSh 100 and costs KSh 70, so the gross profit (the money left after paying for the goods, before rent and your own time) is KSh 30. That is a 30% margin. Now run buy-9-get-1: the customer pays for nine and takes ten.

	Normal (10 sales)	With the card (pay 9, take 10)
Money in (revenue)	KSh 1,000	KSh 900
Cost of the 10 goods	KSh 700	KSh 700
Gross profit on those goods	KSh 300	KSh 200

Illustrative gross-on-goods figures for a 30% margin. The free tenth item is a 10% price cut (one free in ten), but it removes a third of the profit: KSh 300 falls to KSh 200.

Look at what happened. The customer got a 10% price cut. Your profit on those goods fell by a third, from KSh 300 to KSh 200. A small-sounding discount, a big bite out of profit. And it gets worse as margins thin. On a 20% margin the same card halves your profit on those sales. On a 15% margin, roughly two-thirds of it is gone. The thinner your margin, the more a free-item deal hurts. (These are illustrative gross margins to show the method; your own figures may differ. Gross profit is before rent, transport and your own wage, which is why net profit is thinner still⁵.)

The same mechanism works in any currency. A Lagos trader selling a charger at ₦2,500 that cost ₦2,000 is on a 20% margin, so buy-9-get-1 halves the gross profit on those ten sales. Different money, identical trap.

---

DECIDE — does a scheme actually fit your numbers?

One-line takeaway: a reward only pays if it brings extra buying you would not otherwise have had, enough to cover the free goods.

Because so much reward flows to buyers you already had, the whole case for a scheme comes down to one word: incremental. Extra buying the reward genuinely caused. Here is the clean break-even.

That free item costs you its cost price, KSh 70. Each genuinely extra item the scheme causes you to sell earns you the margin, KSh 30. So to pay back one free unit you need KSh 70 ÷ KSh 30 ≈ 2.3 extra sales per completed card. If a typical card-holder would have bought those nine anyway, the card did not cause 2.3 extra sales. It simply gave away about KSh 70 of profit per card.

So the decision is not “should I run a loyalty scheme?” It is “will this reward bring me enough genuinely new buying to cover what I give away?” For a shop full of regulars who already come every week, the honest answer is often no: you would be paying people to do what they already do. A reward is a multiplier on loyalty a customer already feels. It is not a substitute for it. If people like your shop, a reward compounds that. If they only tolerate it, a reward just subsidises them until someone offers a better one.

An honest note on the local picture. Solid published numbers on loyalty schemes for small, owner-run African shops are genuinely thin. Most research covers large, formal companies, telecoms and online retailers^6,7. The mechanism above still applies. A precise “this scheme lifts repeat buying by X%” figure for your kind of shop does not exist in any trustworthy form, so be wary of anyone who quotes one.

---

ACT — design a reward that protects your margin, or skip the discount

One-line takeaway: if you run a reward, reward the extra behaviour, fund it from a fat-margin item, or give something valued that does not cut your price.

If you decide a reward fits, a few moves keep it from eating you alive:

• Reward extra buying, not the buying you already had. Tie the reward to a bigger basket, a slow day, or a higher-margin line, so it pulls behaviour you would not otherwise have got, instead of discounting your steady regulars. For example, a discount on a normally slow time, say Tuesday mornings, to pull sales into it can be smart; the very same discount during your busiest time, a weekend when sales already peak, may just hand profit to buying you had anyway. So talk it over with your advisors and watch your own local realities before you set the terms.
• Fund the reward from a fat-margin item. Give the free thing from a line where the margin is generous, so the percentage bite is smaller than on a thin fast-mover. For example, make the free item a well-marked-up phone accessory or a sachet of spice, not a free bag of sugar or a soft drink you barely make anything on.
• Prefer a reward that is not a price cut. A genuinely valued extra that costs you less than it is worth to the customer, priority access, a small useful add-on, a real service, protects your price and builds actual preference instead of training people to expect discounts. For example, free delivery on a large order, letting a good customer skip the queue at your busiest hour, or bundling in a quick set-up or repair, can be worth far more to them than the little it costs you. Training customers to buy only on promo quietly wrecks the price they think is normal, which is its own slow leak (see the psychology of pricing; and the same logic applies to unmanaged discounting more broadly, one of the quiet profit leaks a healthy business watches for).
• If you use a card, start it “already begun.” Hand the card over with the first one or two stamps already on it, rather than a blank one. In a careful field experiment, a card that came with the first stamps already filled was completed far more often than an identical card that started empty, 34% against 19%⁸. The head-start feels like progress worth finishing. It is an honest nudge. Just remember it moves completion of a purchase, the shallow kind of loyalty again, not deep preference.

And sometimes the best move is to skip the discount and spend that same margin on the thing rewards cannot buy: being a shop people genuinely prefer. That is what turns a stamped card into loyalty that actually holds (see winning and keeping loyal customers).

---

MEASURE — did the reward cause extra buying, or just give profit away?

One-line takeaway: watch incremental buying against what you gave away, not how many cards got stamped.

Redeemed cards are the wrong scoreboard, because they count buying you may already have had. Watch two things instead:

• What you gave away. Roughly, free goods handed out times their cost price. That is the real bill for the scheme.
• Whether buying actually rose. Compare repeat-purchase rate or basket size for card-holders against everyone else, over a few months. Be honest that this is a rough signal, not proof: the people who take cards may already be your keenest buyers. Still, if their buying is not visibly higher than the rest, the scheme is probably paying for behaviour you already had.

If what you gave away is bigger than the extra profit you can actually see, the scheme is a leak wearing the costume of loyalty. When you want the fuller picture, of which this is one piece, that is what a business health check is for.

Find out how healthy your own business is

Find out how healthy your own business is. Reading about the stamp card is one thing. Knowing whether your rewards are winning new business or quietly giving profit away is far more useful. The Monyvo Business Health Check walks you through that honest look, including whether the goods and offers you push are actually the ones that pay. Take the free Business Health Check. If you want to see the margin maths on your own prices first, the Markup & Margin Calculator works it out in a minute.

Related reading

• Winning and keeping loyal customers — the bigger picture on real loyalty, which rewards can amplify but never manufacture.
• Five invisible profit leaks — unmanaged discounting is one of the quiet leaks, and a free-item card is a discount in disguise.
• The psychology of pricing — why training customers to buy only on promo erodes the price they think is normal.

Evidence & Confidence

How sure are we of the main ideas here? Rated by the basis of each claim, not by how much we like it.

Claim	Confidence	Basis
Loyalty has two parts (repeat buying vs genuine preference); rewards mostly touch the first	★★★★☆	Seminal marketing framework, widely used; definitional, not a measured effect
Loyalty schemes tend to raise costs without creating much extra loyalty, because much reward goes to buyers you already had	★★★★☆	Consistent, sceptical peer-reviewed record
A buy-9-get-1 card is ~a 10% price cut that removes a much larger share of gross profit	★★★★★	Transparent arithmetic, shown in full, on illustrative African gross margins
A pre-filled (“head-start”) card gets completed more often than an empty one	★★★★☆	One careful field experiment; moves buying behaviour, not proof of deeper preference
How much a stamp card lifts repeat buying in a small African shop	direction only	Direction is clear (small, often among already-frequent buyers); no trustworthy magnitude exists, so we publish none
“Loyalty members spend X% more” / “retention up X%” style figures	not published	No traceable primary applies to a small shop; these are folklore, so we teach the mechanism instead

Sources

1. 1. Dowling, G. & Uncles, M., “Do Customer Loyalty Programs Really Work?,” MIT Sloan Management Review, 1997. Schemes tend not to alter the market and largely raise marketing cost without creating extra brand loyalty.
2. 2. Uncles, M., Dowling, G. & Hammond, K., “Customer Loyalty and Customer Loyalty Programs,” Journal of Consumer Marketing, 2003. In competitive repeat-purchase markets, buying is driven more by habit than strong preference, so programme potential is limited.
3. 3. Dick, A. & Basu, K., “Customer Loyalty: Toward an Integrated Conceptual Framework,” Journal of the Academy of Marketing Science, 1994. Defines loyalty as the link between genuine preference and repeat buying, giving the true / spurious loyalty distinction. Conceptual framework, used here for the definition only.
4. 4. Sharp, B. & Sharp, A., “Loyalty Programs and Their Impact on Repeat-Purchase Loyalty Patterns,” International Journal of Research in Marketing, 1997. A real programme, measured against normal buying patterns, showed only weak extra loyalty.
5. 5. African small-business margins (verified dataset: Small Firm Diaries; FSD Kenya; StatsSA). Margin ranges used for the worked arithmetic. The examples are illustrative gross-on-goods figures; they are not net profit, and owner labour, rent and transport are separate.
6. 6. Roberts-Lombard, M. & Jaiyeoba, O., South African Journal of Economic and Management Sciences, 2025. Among South African telecom customers, trust drove intention to stay and buy again. Formal-sector context, not small owner-run retail.
7. 7. Africa Loyalty Programs Market Databook, 2025 (commercial market report). Illustrative context that African firms actively pursue loyalty and cashback schemes; no numeric claim rests on it.
8. 8. Nunes, J. & Drèze, X., “The Endowed Progress Effect,” Journal of Consumer Research, 2006. In a car-wash field experiment, a card pre-filled with two stamps was completed 34% of the time against 19% for an identical empty card needing the same number of purchases. Moves purchase completion (behavioural), not proof of deeper attachment.

Try these free Monyvo tools

• Markup & Margin Calculator — see in a minute what a free-item card really costs your own margin.
• Business Health Check — an honest look at whether your offers and rewards are actually the ones that pay.