Payoff Reads · Foodservice

Is our food program winning repeat visits and bigger baskets, or just adding cost?

Find out whether your food program pays for itself after food costs, waste and labor, and how it changed what customers do: how many drivers come inside from the pump, how often regulars come back and how much they spend. Each store is read vs. a twin built from stores that hadn't launched it yet.

15 Pays for itself
6 No significant effect
3 Costs money

Heartland Markets is a fictional chain. Every number on this page is made up to show the format.

What your numbers can't show

Today's food sales reports show what the kitchen sold. They can't show the incremental business it brought to your store.

What you have: a weekly report of food sales, with every store aggregated into one number.
What it can't show: which stores the program is working at, or how much of that food would have sold without implementing the new program. Some of the pizza replaced a hot dog or a sandwich the customer would otherwise have bought, which needs no oven and less labor, and some stores' sales rose for reasons that had nothing to do with food. It also can't show whether the program changed what customers do: how many drivers come inside from the pump, how often regulars come back and how much they spend.
What the read shows: the incremental value the program brought each store: more drivers coming inside, more repeat visits, bigger baskets, and the margin left after food costs, waste and the added labor.

In 2025, food was 28.5% of convenience stores' in-store sales and 38.9% of their in-store gross profit (NACS). Gross profit leaves out the labor a kitchen takes, so food can look like the most profitable thing in a store where the food program actually loses money. That's why our read takes out the labor and the waste.

Watch a pizza rollout get read, store by store

Silent film: how a read finds the stores where a pizza launch worked, and why it didn't at the rest.

What lands on your desk

A food program, as Seurat reads it

Heartland Markets launched a pizza program at 24 of its 42 stores between March and June 2025, 1 or 2 stores a week as ovens arrived. By September, pizza was selling $43,200 a week across the 24. Each pizza store was read vs. a twin built from stores that didn't have pizza yet, including the 18 still waiting for it.

The answer, in three lines:

Customers changed what they do: more fuel customers came inside (+0.8 points), Heartland Rewards members came back more often (+0.3 visits a month), and dinner baskets grew (+$0.70).
The pizza program pays for itself: after food costs, waste, the added labor and the food it replaced, each pizza store adds $340 a week to the bottom line, and that's statistically significant.
It doesn't pay at the 2 highway stores. The share of drivers who came inside didn't move there, so the oven's labor and waste went unpaid.

per pizza store, per week, vs. stores that didn't have pizza yet

What changed	Result	Verdict
Inside visits	+95	Significant lift
Share of fuel customers who came inside	+0.8 points	Significant lift
Visits per Heartland Rewards member, per month	+0.3	Significant lift
Dinner basket, 4 to 8 p.m.	+$0.70	Significant lift
Margin on pizza, after food costs and waste	+$810	Significant lift
Margin on other food the pizza replaced	−$140	Significant drop
Margin on drinks and snacks customers bought with it	+$190	Significant lift
Added labor for the oven and prep	−$520	Cost of the program
What the program adds to the bottom line	+$340	Pays for itself

Sales vs. the bottom line: pizza added $1,300 a week to Store 12's food sales, the number in the film. After food costs, waste, the added labor and the hot dogs and sandwiches it replaced, it adds $380 a week to the bottom line.

Attachment: 41% of pizza orders came with a fountain or bottled drink, which is where most of the +$190 in drinks and snacks came from.

store by store (6 of 24)

Store	Pizza since	Program adds, per week	Verdict
Store 25, Maple Dr	Mar 2025	+$720	Pays for itself
Store 12, Ely Rd	Apr 2025	+$380	Pays for itself
Store 31, Cedar St	May 2025	+$180	No significant effect
Store 9, Boyson Rd	Apr 2025	−$90	No significant effect
Store 27, Pine St	May 2025	−$460	Costs money
Store 40, Hwy 151	Jun 2025	−$650	Costs money

By store type: pizza pays at town stores, where people stop on the way home for dinner. At the 2 highway stores, it costs money: the share of drivers who came inside didn't move, and most of them buy fuel and leave before dinner.

by store type, per pizza store per week

Store type	Program adds	Confidence band	Verdict
2 highway stores	−$610		Costs money
The other 22 pizza stores	+$425		Pays for itself

By daypart: most of the added visits came between 4 and 8 p.m. Lunch pizza sold too, but mostly to customers who'd otherwise have bought a hot dog or a sandwich, which need no oven or extra staff, so at lunch pizza added cost without adding sales.

The limited-time offer: in September, the chain ran a buffalo chicken pizza for 6 weeks at all 24 pizza stores. It sold, but mostly to customers who were already buying pizza. It didn't bring anyone new in while it ran or after it ended, and after its waste it about broke even.

the 6-week buffalo chicken pizza, per store per week

What changed	Result	Verdict
Buffalo chicken pizza sales	+$420	Significant lift
Regular pizza sales it took	−$260	Significant drop
Food margin while it ran, after its waste	+$50	No significant effect
Visits while it ran	+10	No significant effect
Visits in the 6 weeks after it ended	0	No significant effect

KeepFixStopWhat to do: keep pizza at the town stores, fix the 3 stores where it never got going (Stores 9, 22 and 27, worth about $3,500 a week in sales), and leave highway stores out of the next wave. Judge the next limited-time offer on margin, not on its own sales.

The confidence band shows the 95% confidence interval. When the band spans zero, the result isn't statistically significant. Heartland Markets is a fictional chain. Every number on this page is made up to show the format.

How the read works

Food programs usually launch a few stores at a time, as ovens and training arrive. That staggered rollout is what makes the read possible: the stores that haven't launched yet show what the launch stores would have sold without the program.

Give each pizza store a twin: a predicted line of weekly sales and visits, built from stores that hadn't launched pizza yet, weighted so it matches the store for a year or more before launch.
Count the opening weeks separately. The first 2 weeks after launch, with their free samples and launch signs, aren't counted as the program.
Follow the food, and everything around it. The read tracks pizza sales, the food the pizza replaced, the drinks and snacks bought with it, how many fuel customers came inside and how often members came back, so the program gets credit only for its own contribution.
Put a price on it, and check it. Each line is valued at your margins, and waste and the added labor come off. Each twin has to have tracked its store closely before launch. A placebo test then reruns the read on a made-up launch date, months before any pizza: with no program to find, it should show no effect. If it shows an effect, the read isn't working properly, and we tell you.

Can your food program be read?

It depends on how the change launched, more than on how many stores you run. There are 3 routes.

1
The program launched store by store
How the read works
Each launch store vs. a twin built from stores that hadn't launched yet
What it answers
What the program adds per store, after food costs, waste and labor
2
A limited-time offer ran at every store at once
How the read works
Each store vs. its own twin, built from its history and the sales the offer shouldn't touch, like fuel
What it answers
What the offer added while it ran, what it took from the regular menu, and whether visits held after it ended
3
Some stores cut or changed the menu
How the read works
Stores that changed it vs. stores that didn't
What it answers
Whether the shorter menu kept its sales and cut waste

We check your launch dates and menu history on the first call, before you commit to anything.

Scope this read

What we need, and what stays with you

We need:

Weekly sales by store, split into food, packaged food and snacks, drinks, and everything else
The date each store launched the program, menu change or offer, and when an offer ended
Weekly food waste by store
Labor hours by store and week
Your margins by category, as percentages
Transactions by hour, with and without fuel, if your system keeps them
Where your loyalty program can follow members, visits by member, with member IDs you've scrambled (hashed) yourself
1 to 2 years of history

We never ask for:

Names, phone numbers, emails or anything else that identifies a customer
Employee names, shifts or payroll records
Your chain's totals or your P&L

How we handle your data →

How long it takes

About 2 weeks from the day your data lands, then a 60-minute readout with whoever needs to hear it.
We scope it to your chain on a short call: which stores, which menu changes, which offers.

Scope this read Book a call

Questions buyers ask us

Our food sales are up. Isn't that enough? Food sales count everything the kitchen sold, including food customers bought in place of something else and visits that would have happened without the program. The read counts only the program's own contribution, after food costs, waste and the added labor, and shows how it changed what customers do.
Our food sales fell the same season we raised prices and launched a loyalty program. Can you tell which one did it? Sometimes. When changes land at the same stores in the same weeks, no method can pull them apart. They can be separated when they reached different stores at different times, or when they show up in different numbers: non-members paid the same higher prices without the program, so comparing members to non-members points to which change did it. In the sample above, the loyalty program launched at all 42 stores the same month pizza started, so the pizza stores' twins had the program too, and it drops out of the pizza read. We check which of these your data allows on the first call.
Can you tell which menu items earn their place? Yes, where items came on or off the menu at different stores or different times. The read shows which items brought in visits and which only took sales from the rest of the menu.
What if the read shows the program doesn't pay? A read with no significant lift is a private finding. You see every result first, and nothing goes anywhere without your approval.
Do we have to connect any systems? No. One export from your back-office system, sent however your IT team prefers.
Who does the work? We do. Your team runs the export, shares it with us, and comes back to hear our readout.

After this read

Read several changes in one request, or keep the program measured week by week.

Scope several changes →How monitoring works →

Sources: NACS, "U.S. Convenience In-Store Sales Top $340 Billion," April 2026.