In brief
An online store knows exactly who every customer is and has nowhere to ask them to join. The program has to live on the order confirmation, the account page and the post-purchase email, and for a small store credit beats points on every measure except one.
The short answer
An online-only business has the opposite problem to a local one. Identity is free, because every order already carries an account, an email address and a full purchase history, but there is no counter, so nothing prompts anyone to join and no moment makes the program feel like it belongs anywhere. Online loyalty programs have to live on the three screens that replace the counter: the order confirmation, the account page and the post-purchase email. For a small store the reward that works on those screens is store credit, because it is bounded, it can only be spent with you, and it needs no exchange rate.
Identity is free, the moment is missing
Everything a shop struggles to capture, an online store already holds. Everything a shop gets for nothing, an online store has to build.
Behind a counter, somebody has to ask, the customer has to remember, and the card has to be produced before the till closes. That is the genuinely hard part of a physical program, and it is the subject of retail loyalty programs beyond another card. Online, checkout collects a name, an email and an address because the parcel cannot leave without them. Enrolment is not a data problem.
What is missing is the prompt. Nobody says “are you collecting stamps?” There is no poster by the till, no person, no queue with something to read on the wall. A program that waits to be discovered will not be, so every part of it has to sit on a screen the customer was already going to look at.
Credit, points, or a paid club
Three instruments, and they are not variations of one thing. They differ on how much work they are, how tightly you can control what they cost, and who they suit.
| Store credit | Points | Paid club | |
|---|---|---|---|
| Effort to set up | Lowest. One balance, one rule | Highest. An earn rate, a burn rate, and a page explaining both | Middle. Billing, plus whatever the benefit is |
| Cost control | Bounded. Costs exactly its face value in margin | Floats. A balance you owe that grows quietly until people redeem | Revenue is predictable, use is not |
| What the customer must understand | “You have £5” | “You have 340 points” | “What do I get for £29 a year” |
| Fits | Small stores with moderate repeat rates | High volume, or earn rates that genuinely vary by product | High frequency, plus a recurring cost worth removing |
Store credit
The cleanest of the three for a small store, and the easiest to price. A £5 credit costs £5 of margin and never a penny more, whatever the customer spends it on, provided you give no change and it applies to one order.
It is expensive by the standards of a cafe giving away eighty pence of coffee, and that comparison is where the reward classes start. Accept it for the same reason a shop does: you do not make anything. You buy finished goods and resell them, so there is no cheap input to be generous with.
Credit has one property neither of the others has. It carries no explanation. Nobody has ever asked what a pound is worth.
Points
Points buy you exactly one thing credit does not: an earn rate that varies by product. If some of your range carries triple the margin of the rest, points let you pay more for the things you want moved, and that is a real advantage.
The price is an exchange rate. Published, effectively permanent, awkward to change, and attached to a balance somebody will email you about. It also creates an obligation that grows while you are not looking, since every unredeemed point is something you owe. The mechanics, including how the rate gets set, are on how loyalty points work.
A paid club
A different instrument altogether. The customer pays up front, usually against shipping, and the club has to save them more than it costs or they cancel.
The arithmetic belongs to them, and they will do it. A £29 annual membership against £3.95 shipping breaks even at eight orders. If your typical member places four, the club costs them money and they will work that out at renewal, whatever the welcome email said.
It also changes the failure mode. Break a free program and somebody shrugs. Break a paid one and they want their money back, because they bought it.
The discount code trap
A loyalty reward delivered as a code the customer types into a box competes with every other code that fits in that box.
The mechanism is visible in your own funnel. An empty discount field on the payment screen is an invitation to open a new tab. A customer who leaves to search for a code can come back with somebody else’s affiliate cookie attached, at which point the reward you designed has cost you the reward and a commission on the same order.
Two fixes, and the first is worth more than the second:
- Apply the reward without a field. If the balance is on the account, show it in the cart and subtract it at checkout. A reward that has to be typed is a reward that reminds people codes exist.
- Make it something a coupon site cannot reproduce. Credit tied to one account, early access to a drop, a size held back, a product not sold separately. A generic 10% code is reproducible by definition, which is what aggregators are for.
If you keep the code field for real campaigns, collapse it behind a link rather than leaving a blank box on the payment step.
Frequency decides whether to run one at all
A business selling something people buy twice a year should not run a points program.
The reasoning is the threshold rule from how loyalty programs work, applied to order data instead of visits. A reward has to arrive while the customer still cares. If someone orders every ten weeks, a five-order threshold is fifty weeks away, and nobody holds a program in their head for a year. A balance that has been forgotten is not changing what anyone buys.
One number decides it, and your store already knows it: the median gap between a customer’s first and second order. If that gap is longer than the reward window you had in mind, the program cannot fire, and no amount of design work fixes it.
When frequency is genuinely low, spend the effort elsewhere:
- Time a message to consumption rather than to a calendar. If the product lasts eight weeks, week seven is when to write.
- Work on first to second order, not on the tenth. For a low-frequency store that is the whole game, and it is the same first-purchase problem covered in loyalty programs for small businesses.
- Consider referral instead of retention. Somebody who buys twice a year may still know five people, and a referral does not depend on their own frequency.
Where the program shows up when there is no counter
Four surfaces, and one placement that decides whether any of them work.
- The order confirmation page. The only screen every buying customer sees, at the moment they are most pleased with the decision. Enrolment belongs here, and so does the new balance.
- The post-purchase email. The one email with a real reason to be opened, because it carries the tracking. Put the balance underneath the tracking, not above it.
- The account page. The only place a customer visits deliberately, and where the balance lives permanently.
- The box. An online store does have one physical surface. A card in the parcel showing the balance, or a code that opens the account page, is the closest thing to a counter a warehouse can offer.
The placement that matters more than all four is the checkout itself. Show the credit before payment, not after. A balance somebody discovers once they have already paid changed nothing about the order, and a balance shown in the cart is the only version of this that behaves like a reason to buy.
Worth saying plainly: NeoLoyal is a counter product, where customers join from a printed QR poster and a member of staff confirms each stamp, so an online-only store has nothing to point it at. If nobody in your business ever stands in front of a customer, the tools that fit are the ones built into the storefront you already run.
Write the reward as one sentence and see if it fits
Put it on the order confirmation with no link to terms underneath it.
“You have £5 to spend on your next order” fits. “Earn 1 point per £1, redeem 100 points for £5 off orders over £30” does not, and that is not a wording problem. It is a program that needs its own page to explain itself, which is a fair thing to build once your order data shows the frequency to support it, and an expensive thing to build before then.