The owner posted “show this story for 15% off”. By Saturday the floor had no way to know who had already used it, who was a regular, and who just flashed a screenshot from a friend. The till had a sticky note: “ask manager”. At 8pm nobody asked. A week later the discount line was up and repeat visits were not. That is not a loyalty program. That is a leak with a social caption.
A restaurant loyalty program works when it rewards a behavior you actually want. A blanket cut for anyone who asks mostly cheapens the bill for people who were coming anyway.
Why “anyone who asks” does not build returns
A discount on its own does not say why they should come back to you, when, or what should change. A café that knocks 15% off every latte for a smile still sees the same morning rush and the same empty mid-afternoon. A new guest uses it once and leaves no handle you can invite. Staff apply the rule when they remember. You see reward cost. You do not see programme effect.
Points, stamps and discounts are not the enemy. They cannot be the whole idea. Loyalty is part of service: recognise the guest, apply a rule at the till or on the order, then check whether the mechanic fits the venue’s rhythm.
Name the return you want before you name the prize
A dining room may want bookings off-peak. A pizzeria may want a repeat on its own channel rather than a platform. A café may want a morning habit or a return for a seasonal pastry. Sushi may want a reminder about a set and collection.
Three answers, or you are just building a prettier notepad:
- who should return,
- when they should return,
- what they get so the reward is fair to the guest and survivable on margin.
If you cannot say that, a new module will issue the same discounts with a terms page.
Match the mechanic to the venue, not to a trend
A café with small, frequent tickets is not a casual dining room and not a collection-led shop. One points card will not fit all three.
| Program model | When it fits | Watch-out |
|---|---|---|
| Points on spend | Guests return often and grasp the rule without a speech | An oversized reward becomes a standing cost |
| Reward after a run of visits | You want to reinforce a habit | You must define what counts as a visit |
| Coupon for a moment | You want a quiet hour or a specific item | Do not cut the peak bill for no reason |
| Offer for regulars | You already know who comes back | Rules must be simple for staff and readable for guests |
The table does not pick a winner. It moves you off “we want loyalty” onto an ops choice. Format differences: OrderNow industries.
Five things that must work before the story goes live
Loyalty exposes process holes fast. If the team does not know when to ask for a phone number, where points show, who can issue a reward, or how to take a coupon, the program only exists when a manager is hovering.
- Recognition. Phone, account, order history: one simple method. Manual typing breeds duplicate people.
- Earning rules. A server cannot explain a policy for three minutes in a queue.
- Visibility at sale. POS, online order, collection, payment. If staff jump apps, the program vanishes at peak.
- Promotion control. “Always, on everything” makes it impossible to see whether returns improved or bills just shrank.
- Contact. Consent and message rules before the first “you have points” text, not a week after launch.
Loyalty at the till, not in a group chat
In OrderNow, loyalty and coupons sit in Pro, not Start. Direct online is also outside Start. Do not buy a retention stack if you only need a calmer floor. If returns, coupons and your own guest list are the priority, pick the plan on purpose.
OrderNow is not a marketplace and does not automatically replace a fiscal device. It is useful when points, sales, menu and coupons are not three notebooks. See the feature catalogue, the loyalty product page and coupons.
Loyalty is a bad first move if the menu is stale, orders are chaotic and nobody knows which channel works. Points will not fix a slow pass or a wrong plate. They can hide the cause for a week.
After the first weeks, do not judge only on rewards issued. Ask whether staff remember the program without nagging, whether guests get the rule, whether it hits the actual goal (off-peak return, own-channel reorder), whether the manager can see redemptions, and whether discounts fire where you do not need them. Sometimes the fix is a simpler rule or splitting loyalty from ordinary promos. Owned-channel context: when an owned sales channel makes sense.
If you already have one mechanic and want to walk it on a till, book a presentation.
Questions and answers
Does a restaurant loyalty program have to be built on discounts?
No. A discount is one reward, not the only one. A product, a regulars’ item, an invite in a quieter hour or a coupon for a specific occasion can work better. The reward should reinforce the behavior the venue needs, not automatically shave every bill.
Do loyalty points replace decent service?
No. Points can remind someone you exist and organise how rewards are handed over. They will not shorten a queue, unmake a wrong plate or make an unreadable menu readable. If guests stay away for those reasons, fix the process first.
Should in-house and online orders share the same program?
They do not have to launch on the same day, but they should not treat the same person as a stranger. Points earned in the room that vanish on your own web order teach the guest the program is accidental. Wider tool context: online ordering software criteria.