Stock with a deadline
Most loyalty programmes buy a return visit. A bakery's has a second job that starts mid-afternoon: everything left in the case loses all of its value at closing. The flour, labour and oven time are already spent. An unsold sourdough is not thinner margin — it is a cost with nothing opposite it.
So a reward here is not what it is elsewhere. One handed over in the last hour is often made of stock heading for the bin: it costs less, and later. Three facts pull against that — small ticket, high frequency, an unforgiving morning peak.
A timed multiplier beats a markdown
The obvious closing tool is a markdown — thirty per cent off after five. Cutting the price beats binning. But it reaches everyone, including the customer who came in ready to pay list, and it leaves margin in cash today. A timed offer multiplier holds the price and lifts the earn rate for a window instead.
| Question | Markdown | Timed multiplier |
|---|---|---|
| Who it reaches | Everyone at the counter | Members only, so it is targeted |
| Cost on a $6 basket | $1.80 of margin | About 28 cents of future goods |
| When you pay | At the till, today | Later, on redemption |
| What you learn | Nothing about who bought | Member, basket and hour |
| What it trains | Come at five, pay less | Come at five, return to spend |
| Where it fails | Erodes the full price | Nothing for a non-member |
The working, assumptions stated. Average basket $6.00. One point per dollar. The reward is 60 points for an item selling at $4.50 that costs $1.40 to make, so a point is worth about 2.3 cents of goods to you. A $6 basket earns 6 points, near enough 14 cents; tripled, 18 points, about 42 cents — 28 cents more. The markdown on the same basket costs $1.80, more than six times as much, in cash rather than pastry.
Cost points as though every one is redeemed; breakage is a windfall, not a plan. And vary the window — an hour that never moves has not cleared waste, it has moved your peak into your cheapest hour.
Run it on the days you actually over-bake, and announce it the same day. Merchant-composed, consented email can queue through an operator-enabled Resend path, SMS or WhatsApp through an operator-enabled Twilio path, and the member page can use web push for people who allow it. Provider status is kept distinct from delivery evidence, while QR campaigns carry a budget cap and a kill switch. Check those in the current feature list first.
Daily bread and occasion cake
Two people walk into a bakery and they are not the same business. One buys a loaf four mornings a week. The other orders a birthday cake in March and a Christmas box in December.
| Signal | Daily-bread customer | Occasion customer |
|---|---|---|
| Visit rhythm | Several times a week | Once or twice a year |
| Ticket | Small and consistent | Large and irregular |
| What motivates | Visible progress toward a near milestone | Value banked against a planned purchase |
| Mechanic that fits | A stamp card | Points |
| What the other does | Points on $6 accrue too slowly to feel like anything | One stamp for a $400 cake reads as an insult |
Rewardfinity ships a bakery card template — the line reads “Little treats add up.” — paired with a stamp card, which takes a required count, stamps per purchase, a minimum spend, up to five milestones and remainder carry.
Tune both to the same funded discount and the choice stops being about cost. A $1.40 reward per $60 of spend is 2.3%. An eight-stamp card whose reward costs $1.10, on the same $6 basket, is $1.10 per $48: also 2.3%. What differs is which customer the mechanic describes, and the fuller comparison of points against stamp cards works through the rest.
Many bakeries need both — stamps at the counter, points as the ledger for pre-ordered work. It goes wrong when both reward the same sale in two currencies, because staff then have two things to explain.
The queue is the constraint
At quarter past eight the limit is not motivation, it is seconds. A question added to a transaction gets dropped by whoever is on the till, quietly, and you find out weeks later from a flat sign-up count. Joining has to be one scan: a QR code at the payment terminal, a member page that opens in the browser, no install. Google Wallet is a configured path; Apple Wallet public launch remains dark pending its physical canary. Confirm the exact state before printing signage. Delivery trade-offs sit on the member card page, counter flows on in-store loyalty.
Give staff one sentence, said after the payment prompt every time. Counter actions are attributed per staff member and can feed a staff-facing spin wheel with an owner-set allowance, so recognition is a mechanic rather than a memo.
Pre-orders and standing orders
The Friday call for Saturday, the office taking two dozen every Monday, the cake collected at a set time: none of them stand in the queue, and several outspend the people who do.
Decide how they earn: a receipt claim, a native point-of-sale or Shopify sale, or a write through the REST API. Then decide whether a large order earns at the counter rate. Points are a percentage, so a $400 cake at a point per dollar mints 400 points: six rewards, about $8.40 of goods. Consistent arithmetic, possibly the wrong answer for someone who appears twice a year, which is what per-order caps are for. Make their reward one that brings them in early — a dozen pastries, not a discount on the cake they were ordering anyway.
Launch checklist
- Set the stamp threshold from the discount you will fund.
- Decide whether pre-orders earn, and cap them if the rate is wrong.
- Pick the multiplier window from your waste log, not a fixed weekday.
- Run one transaction end to end, including a refund, before staff see it.
- Set the review date and what result would change the threshold.
Plans differ by country, so check what your economy's price includes first.
What bakery owners ask before launching
What is the best loyalty program for a bakery?
A stamp card for the everyday counter purchase, points for pre-orders and cakes, and a timed points multiplier in the last hour to move perishable stock. Stamps fit a small repeatable basket. Points fit large irregular orders, where one stamp against a $400 cake looks absurd.
How do I sell bread at the end of the day without discounting?
Raise the earn rate instead of cutting the price. A triple-point window pays in future stock rather than cash, and costs you only on redemption. On a $6 basket with a point worth about 2.3 cents of goods, tripling costs roughly 28 cents against $1.80 for a markdown.
Should a bakery use a stamp card or points?
Often both, for different customers. Stamps suit the person buying a loaf four mornings a week: progress is visible and the milestone close. Points suit the person ordering a cake twice a year, because a percentage records what a large order is worth. Keep one mechanic per purchase type.
Does a loyalty program slow down the morning queue?
It does if joining takes more than one scan. Put a QR code at the payment terminal, use a member page that opens in the browser with no app install, and give staff a single sentence to say. An ask that needs explaining quietly stops happening in a rush.
Can I give loyalty points on a cake pre-order?
Yes, if the order is recorded somewhere the programme can read: a receipt claim, a native point-of-sale or Shopify sale, or a write through the REST API. Decide first whether a large order earns at the counter rate, because points are a percentage and a $400 order can mint several rewards at once.
How much does a bakery loyalty program cost to run?
Two separate costs. The software is a subscription, with a free plan covering up to 100 members and one stamp card, and paid plans priced to local purchasing power rather than one global figure. The larger cost is the reward itself — cost of goods on every completed card.
Cost the reward, then pick the window.
Start with one stamp card and a waste log. Add the multiplier once people have joined.