Creating the Best Loyalty Program for Your Restaurant

Ask 10 restaurant brands why they're launching a loyalty program, and eight of them will give me the same answer: because everyone else has one.

Loyalty has become the status quo of restaurant marketing. Everyone thinks they need a program. But that status quo doesn't hold up once you ask a few follow-up questions. A guest who's already ordering three times a week doesn't need an incentive to keep doing what they're already doing, and offering one anyway can cost a brand money it didn't need to spend.

I've spent years helping restaurant brands untangle the difference between a loyalty program that builds habits and one that just eats away at margin. Here's what I've learned.

Why do restaurants need a loyalty program in the first place?

Before covering points, tiers, or tech stack with a brand, I ask one question: why?

“Because everyone else does” is not a reason. A thoughtful answer sounds more like: 50% of our guests don’t return for a third visit, which means we never pay back the cost to acquire them, and we think the right program could solve our retention problem. That's a sound starting point and a foundation for building loyalty.

Once a brand has a data-backed reason, we dig deeper: what do you actually know about guest behavior beyond their average visits? Not all guests are created equal and, frankly, not all of them need to be the focus of a loyalty program. A guest with high lifetime value who has already formed a habit with you doesn't need a discount to keep doing it. Give them one anyway, and the program has just cut into margin it didn't need to touch.

My rule for that group is simple: I leave them alone.

If we treated loyalty programs like we treat human relationships, we'd design them differently. Most brands overmarket to guests who are already good to them (you don't need to prove anything to a true friend), and underinvest in the ones they actually need to nurture (if you never check in with a new acquaintance, or invite them to do something you think they’ll enjoy, how do you build a friendship?).

The guests worth building a program for are the ones who show up sometimes, not the ones already showing up every week. Of course, we value those guests and want to treat them well, but we don’t want to create an addiction to discounts.

What should you know about your menu before you build a loyalty program?

Before I sketch out a structure, I want to know what's driving repeat visits, and it usually has little to do with how good your email marketing is.

I worked with one brand that found that guests whose first order was the Greek yogurt came back at roughly three times the rate of guests who ordered a salad (unexpected, because this was a “salad” company, so to speak). We asked ourselves, “What if we used the Greek yogurt as our first-order loyalty incentive? Would we drive repeat?” We thought so, and spoiler alert: we ended up giving every rewards member a free yogurt when they signed up (no strings attached), and the average repeat rate skyrocketed.

Point being, understanding menu performance is key before you define what the program is. That means knowing which items drive a first purchase, which ones bring guests back, and which ones (limited-time offers aside) actually push guests away—Olo Guest Intelligence was built to help with this. Those answers become the foundation of a loyalty program, not the program itself.

Marketers often start with the idea that you just need to throw discounts and promos at problems, but in reality, the heartbeat of loyalty is good food and good service.

How can you test a loyalty concept before building it?

My favorite loyalty tool is about as low-tech as it gets: a punch card.

I think the punch card is a lost art that's overdue for a comeback (maybe not permanently, but worth a test!). Hand some guests a card that requires ten coffees to earn a free coffee. Hand others one that requires five breakfast pastries to earn a free breakfast pastry. Run it for two months, then count which cards actually got cashed in the most.

The logic is about reversibility. A punch card is easy to walk away from if it doesn't work. A points program is not. It's not simple to tell a guest "you've earned 5,000 points, and we've promised you all this value in the bank," and then have to unwind that promise later.

Before you choose between points, tiers, or surprise-and-delight, test what guests actually respond to using the cheapest, most reversible version you can find. Then build the expensive version around what the test told you, not the other way around.

What metrics prove your loyalty program is actually working?

To start, are people even signing up? How do sign-ups compare to your enrollment target? If sign-up is low, ask why. Pro tip: Guests don’t love handing over their name, phone number, email, age, and birthday in exchange for a dollar off a cheeseburger.

After that, look at reward engagement rate and understand what types of rewards actually change behavior. Are high spenders banking points toward a higher-value item, or just hungry for a small discount on every order? It all comes back to whether the value exchange is right.

You’ll learn a lot more if you think of loyalty like scientists think of experiments. They don’t assume. They suspect, test, and learn.

Think of airline loyalty programs: if you fly enough miles, you eventually get a high-value reward (hello, business class upgrades!). That kind of emotional connection is hard to build when the whole model asks a guest to spend $500 to get $10 off. A better strategy might be: "Today, all cheeseburgers are free when you buy a shake" as a surprise-and-delight instead. This leverages a top-converting menu item (cheeseburger) as a reward to get guests to try an item (shake) with a high satisfaction rate that increases average check value by $3.

I'd rather see a brand try something with real emotional weight: a loyalty perk built around unsold inventory at the end of the day, framed around reducing food waste, instead of another dollar-for-dollar tier. You have to think of ways to connect that aren't so obvious. Loyalty should build brand equity, not cheapen it.

As for the metrics that get thrown around most, guest lifetime value and average order value, I'm skeptical of both as loyalty indicators. Lifetime value is circumstantial, and average order value swings on factors unrelated to loyalty. My preferred number is more basic: guest yearly spend. Comparing this for members against non-members is key. If your members are spending $50 less annually on average than your non-members, is something wrong, or is that just the cost of doing (loyalty) business? How do you figure that out? That’s a whole other blog post for me to write in the future.

What's the one thing every restaurant operator should know about loyalty?

My closing advice doubles as a warning against loyalty's most common trap: Frictionless signup paired with poor strategy.

The space is saturated, and most brands respond by making sign-up as frictionless as possible. That's great for enrollment numbers, but without a solid strategy to retain guests long-term, it can drive exactly the behavior you don't want: the guest who grabs a discount on their first order and never comes back. Loyalty programs may only be good at driving that first purchase, depending on how you structure them, and the rest of your brand and marketing has to carry the weight from there.

My last piece of advice: keep data that informs strategy relevant and time-bound. Guest behavior from two years ago won't predict what a guest wants today. Don't build a loyalty program around data that's stale or even worse, no data at all. Use what's most recent, and I say, don't go back further than a year.

Again, the most important question to ask before getting in too deep is: Why? I want restaurants to use loyalty. I just want them to focus on themselves and their guests, not what everyone else is doing.