The number behind the headline
Fon Khunsamart runs Baker St. Cafe, a Thai restaurant and bubble tea shop in McMinnville, Oregon, with her husband and two of her kids helping out. When a customer hasn’t visited in a while, she sends a "we miss you" push notification with a discount — and roughly one in nine people redeem it. It’s a small, human-scale example of a pattern that shows up at much larger scale in Square’s transaction data: customers who visit a business at least four times a year — Square calls them "regulars" — generate about six times the annual revenue of one-time visitors, tip 11 percent more, and tend to buy the same thing over and over. In food and beverage, that’s true of roughly 59 percent of orders; in beauty businesses, 78 percent.
Square’s broader 2026 Local Economy Report adds a second layer to this: regulars aren’t just loyal to one shop, they’re often shared across a neighborhood. About 32 percent of regular customers in a given ZIP code patronize multiple nearby businesses, and each of those "network connections" is associated with meaningfully higher annual revenue — a few hundred to a couple thousand dollars depending on the city. A similar UK-focused analysis from Square found regulars generating at least five times as much revenue as transient customers nationally, rising to ten times in Sheffield, with the effect driven less by bigger purchases and more by sheer frequency — regulars return around eleven times a year on average.
These are striking numbers. They’re also, on their own, a slogan dressed up as an insight — until you ask what a business is actually supposed to do with them.
What the data can and can’t tell you
Here’s the trap: it’s tempting to read "regulars drive 6x more revenue" and conclude that if you just get more people to come back, revenue will multiply. But repeat visits and higher revenue could easily be two symptoms of the same underlying cause — good food, a convenient location, fair prices — rather than one causing the other. A customer becomes a regular because something about the business already works for them; the data doesn’t tell you which something.
The same caution applies to a headline figure in both reports: businesses using marketing tools like email, text campaigns, or loyalty programs were far more likely to retain regulars than those that weren’t — 90 percent versus 38 percent in the U.S. data, a similarly wide gap in the UK version. That’s a real, sizable association. It is not proof that installing a loyalty program will produce that outcome for your bakery or your garage. Businesses that adopt marketing software in the first place may already be more organized, better capitalized, or more customer-focused — the tool could be a marker of a business that was already good at retention, not the cause of it.
It’s worth remembering, too, that these figures come from one payments platform’s own merchant base, surveyed at one moment in time. A six-times multiple in Atlanta, or a ten-times multiple in Sheffield, describes what happened among the businesses Square could see — not a target every shop should expect to hit, and not a universal law of local commerce.
None of this means the underlying instinct is wrong. It means the instinct — "our regulars matter" — is a hypothesis, not a conclusion. And hypotheses are testable.
| What the data suggests | What it doesn’t establish |
|---|---|
| Regulars generate substantially more revenue than one-time visitors | Whether that gap holds for your specific business, category, or city |
| Marketing-tool users retain more regulars | Whether the tool caused the retention, or reflects an already-strong business |
| Regulars are often shared across nearby businesses | Whether partnering with neighbors would actually shift your own repeat rate |
| Product consistency correlates with repeat purchases | Which specific offer, message, or service change would convert a one-time buyer |
| Frequent, small purchases compound into outsized annual value | Whether raising visit frequency is easier or cheaper than raising order size |
From a belief to a testable hypothesis
A hypothesis, in the plainest sense, is a statement that links a specific change to an expected customer response — something you can be proven wrong about. "Our regulars are valuable" isn’t one; it’s a belief. "If we text lapsed customers a small discount within 30 days of their last visit, at least 10 percent will return within two weeks" is a hypothesis. It has a trigger, a timeframe, and a number attached, which means you can find out whether it’s true.
This is where the metric you choose matters more than it seems. Total foot traffic or new sign-ups can look great while telling you almost nothing about whether anyone is actually sticking around. Repeat purchase rate — the share of customers who come back within a defined window — is a much sharper early signal, because it filters out the noise of one-off curiosity visits and shows you who’s forming a habit. It’s also worth remembering that "regular" is relative: someone can visit your coffee shop every Tuesday and never set foot in the bakery next door twice. Loyalty isn’t a trait a customer carries around — it’s a relationship specific to each business, which is exactly why it can be nudged.
flowchart TD A[Observed behavior: some customers return often] --> B[Assumption: regulars drive our growth] B --> C[Hypothesis: a specific change will increase return visits] C --> D[Small test: message, offer, or service tweak] D --> E[Measured decision: repeat visits and spend before vs after]
Running the test without overclaiming
The good news is that testing a retention hypothesis doesn’t require a data science team. It requires discipline about what counts as evidence.
Start with observation: pull your own point-of-sale records and see who actually returns four or more times a year, and how much they spend compared to everyone else. This is your baseline, and it may look nothing like a national average — that’s fine, it’s the number you’re actually trying to move.
Then talk to a handful of regulars directly. Ask what makes them come back, and what would make them stop. Simple interviews, five or ten of them, often surface the real driver faster than any dashboard — maybe it’s the same barista every morning, maybe it’s a menu item nobody else has. Khunsamart’s "we miss you" message works partly because it’s personal enough to feel noticed, not because push notifications are inherently magic.
Finally, run one change at a time and measure it against your own baseline: a lapsed-customer offer, a loyalty perk, a tweak to service speed. Track repeat visits and average spend for a defined period before and after. If the number moves, you’ve learned something real about your customers — not proven a universal law, just found one lever that works in your shop, with your customers, right now.
The takeaway
Regular customers are clearly worth paying attention to — the data behind that instinct is broad enough to trust as a signal. But the six-times figure isn’t a target to chase; it’s an invitation to ask a sharper question. What, exactly, is making someone come back a fourth time — and can you find out cheaply enough to be wrong without it costing you much? That’s not a rejection of intuition. It’s how good intuition earns its keep.
