Payoff Reads · Online ordering

How does ordering through an app change your baskets and your margin?

Find out what your own app and delivery apps like DoorDash added or took away at every store. Each store is read vs. a twin built from stores that hadn't launched yet.

Orders: which were new, and which only switched from the counter.
Baskets: what changed when customers ordered from outside the store.
Margin: what's left after app fees and delivery commissions.

Silent film: how a read splits new app sales from orders that only switched from the counter.

What your numbers can't show

Your app and delivery reports count the orders. They can't show how ordering through an app changes what customers buy, or what it does to your margin.

What you have: online orders and sales by store, from your app, your website and each delivery app, and how fast they're growing.
What it can't show: how the apps change ordering at each store. A customer who used to order at the counter and now taps an app is the same sale, now with an app fee on it, or a delivery app's commission of 15% to 30%. Your reports also can't show whether the app's suggestions add items, how often customers just repeat their last order, or what pickup customers stop buying when they skip the aisle.
What the read shows: what online ordering added or took away at every store after its fees: new visits, the counter orders it replaced, what changed in the basket, and the margin left once app fees, delivery commissions and the labor to pack orders come off.

From our national survey

21%of fast-food and c-store customers often pick up orders they placed in a chain's app.
16%often order through a delivery app.
41%of app and delivery users pick or avoid specific locations when they order.
39%of people who switched to ordering online did it to get away from something about the store: the wait, the staff or the store's condition.

Ordering online doesn't hide a weak store.

What lands on your desk

Online ordering, as Seurat reads it

11 Adds to margin
14 No significant effect
5 Cuts into margin

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

Heartland Markets turned on ordering ahead in its own app, for pickup, at 30 of its 42 stores between January and June 2024, a few stores a week as each kitchen was trained. The other 12 don't take orders ahead. In April 2025, it put 10 of the 30 on a delivery app. Each store was read vs. a twin built from stores that hadn't launched yet.

The answer, in three lines:

Heartland's own app added margin where the line was long. About 2 in 5 app orders were new sales, from customers coming more often at the morning and lunch rush. The other 3 in 5 had switched from the counter.
Ordering from outside the store changed the basket: the add-on the app suggests raised food sales, but pickup customers skipped the aisle and bought fewer drinks and snacks.
The delivery app cut into margin: 3 in 4 delivery dollars were sales Heartland would have made at the counter or in its own app, now paying a 25% commission. Delivery cost $110 a week per store.

Heartland's own app, per store per week, first year after launch, vs. stores that hadn't launched yet

What changed	Result	Verdict
Orders placed in the app	110	From the app's records
Visits, counter and app together	+48	Significant lift
Visits, 11 a.m. to 2 p.m.	+30	Significant lift
Add-ons with food orders	+$70	Significant lift
Drinks and snacks bought in the aisle	−$40	Significant drop
Sales	+$520	Significant lift
Margin on the added sales	+$270	Significant lift
App fees and the labor to pack orders	−$90	Cost of ordering ahead
Margin left after fees	+$180	Adds to margin

Sales vs. margin: the app took about $1,050 a week in orders at the average store, and $520 of it was new sales. The rest would have been rung up at the counter. After the app's fees and the labor to pack orders, $180 a week is left in margin, about $9,400 a year.

Where the new orders came from: Heartland's own app mostly reaches customers who already shop there, so its new sales came from them coming more often, mostly at the morning and lunch rush, when the line was longest. Those are customers who used to give up on the line, and now order ahead and skip it.

What's in the basket: ordering from outside the store changed what customers bought. The app suggests an add-on with every food order, and add-ons rose $70 a week per store. Pickup customers skipped the aisle on the way out and bought $40 a week less in drinks and snacks. In the app's records, 4 in 10 app orders repeated the customer's last order, so a new menu item only reaches app customers when the app puts it in front of them.

store by store (6 of 30), margin left after fees

Store	Ordering since	Margin per week	Verdict
Store 33, Kennedy Pkwy	Jan 2024	+$410	Adds to margin
Store 12, Ely Rd	Feb 2024	+$300	Adds to margin
Store 18, C Ave	Mar 2024	+$110	No significant effect
Store 2, Oak Ave	Apr 2024	−$30	No significant effect
Store 23, Hwy 6 E	May 2024	−$150	Cuts into margin
Store 40, Hwy 151	Jun 2024	−$170	Cuts into margin

By store type: the app adds margin at the 14 stores near offices, schools and commuter routes, where the morning or lunch line used to send customers away. At the 5 highway stores, app orders mostly replaced counter orders, so the fees and packing labor cut into margin.

by store type, per store per week, margin left after fees

Store type	Margin per week	Verdict
14 stores near offices, schools and commuter routes	+$370	Adds to margin
11 town stores	+$90	No significant effect
5 highway stores	−$150	Cuts into margin

The delivery app: in its first 6 months, the delivery app brought in 38 orders a week per store, about $610 in sales. But total sales at those stores rose only $150 a week vs. their twins, because sales at the counter and in Heartland's own app fell by $460. So about 3 in 4 delivery dollars were sales Heartland would have made without the delivery app, and the store pays the 25% commission on all of them. Delivery cost $110 a week per store, and it cut into margin at 7 of the 10.

the delivery app at 10 stores, per store per week, first 6 months, vs. app-ordering stores without it

What changed	Result	Verdict
Orders through the delivery app	38	From the delivery app's records
Sales through the delivery app	$610	From the delivery app's records
Sales at the counter and in Heartland's own app	−$460	Significant drop
Sales, all channels together	+$150	No significant effect
Margin on the added sales	+$60	No significant effect
Delivery commission, 25% of every delivery order	−$150	Cost of delivery
Labor to pack delivery orders	−$20	Cost of delivery
Margin left after commission	−$110	Cuts into margin

KeepFixStopWhat to do: keep app pickup at the 14 stores where it adds margin, and promote it at the counter during the rush. Fix the town stores where it breaks even by putting the pickup shelf by the door, so app customers skip the line. Then win delivery customers over to Heartland's own app with an offer in every delivery bag: they skip the delivery app's fees, and Heartland pays a small app fee instead of a 25% commission. Stop the delivery app at the 7 stores where it cuts into margin, or raise delivery prices there until the commission is covered.

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.

Try it on your stores

See how this works at your stores, with a free Fit Check

This page explains the method behind a read. A Fit Check shows it in practice, on your own stores: pick a concept we've seen working in real stores, and see which of your stores look like the places where it works, and where to try it first. It's free, you don't send any sales data, and the quick read takes about a minute.

Order aheadWhere to try ordering ahead first, store by store, each with a twin.

How the read works

Online ordering usually turns on a few stores at a time, as kitchens are trained, and delivery apps add stores in waves. That staggered start is what makes the read possible: the stores that hadn't launched yet show what the launch stores would have sold without online ordering.

Give each store a twin: a predicted sales line, with visits by hour, built from stores that hadn't launched online ordering yet, weighted so it matches the store for a year or more before launch.
Count every channel together. Most POS exports record each order's channel: the counter, your app, your website or a delivery app. The read adds them all up, so an order that only switched channels doesn't count as new.
Read the hours and the basket. Online orders bunch up at the morning and lunch rush, so the read looks at visits hour by hour. It also looks at what changed in the basket: the add-ons the app suggests, how often customers repeat their last order, and the drinks and snacks pickup customers no longer grab from the aisle.
Evaluate what it's worth, then test the read. Each line is valued at your margins, and app fees, delivery commissions and an estimate of the labor to pack orders 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 online orders. Because there's nothing to find during this period, the placebo test should show no effect. If it shows an effect, the read isn't working properly, and we tell you.

Can your online ordering be read?

It depends on how ordering reached your stores, more than on how many stores you run. There are 3 routes.

1
Ordering turned on store by store
How the read works
Each store vs. a twin built from stores that hadn't launched yet
What it answers
Which orders were new, what changed in the basket, and the margin left after fees
2
A delivery app at some stores
How the read works
Stores on the delivery app vs. a twin built from stores without it
What it answers
Whether delivery added sales, or switched them to a channel with a commission
3
Ordering launched at every store at once
How the read works
Each store vs. its own twin, built from its history, from sales ordering shouldn't touch, like fuel, and from stores where far fewer customers order online
What it answers
What ordering added or took away at each store, and at which hours

We check your launch dates and how your orders are recorded on the first call, before you commit to anything.

Scope this read

What we need, and what stays with you

We need:

Sales by store and week, split into food, drinks and everything else, with each order's channel and hour, which most POS exports include
Your app's order export, and each delivery app's, by store and week
The date each store started taking orders in your app, on your website and on each delivery app
Your app fees and delivery commissions, and your margins by category, as percentages
1 to 2 years of history

We never ask for:

Names, phone numbers, emails or anything else that identifies a customer
Delivery addresses
Logins to your app or your delivery accounts

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 channels, which launch dates.

Scope this read Book a call

Questions buyers ask us

Our online orders keep growing. Isn't that enough? Not on its own. Online orders count every order placed in an app, including orders customers would have placed at the counter. The read counts only the sales the store wouldn't have made without online ordering, and takes off what the fees cost.
Our POS already shows which orders came through each app. Why do we need a twin? The channel on each order shows what each channel sold, and the read uses it to compare baskets. But it can't show which orders are new. Customers who order delivery aren't a random group: they may also be your biggest spenders at the counter, so a big delivery basket doesn't prove delivery made it bigger. Comparing each store's total sales with its twin shows what the store gained in all, so an order that only switched channels doesn't count as new.
Delivery orders don't come through our POS. Can you still read them? Yes. Each delivery app's own reporting can be exported by store and week. We line it up with your POS sales, so the read sees every channel.
Do delivery apps charge for pickup orders too? Usually, yes, at a lower rate. DoorDash's standard plans, for example, charge 6% on pickup orders, vs. 15% to 30% on delivery. The read takes off whatever your contract charges, channel by channel.
What about kiosks? Kiosks take orders inside the store, so they change different things: the line, and the add-on the screen suggests. They're read the same way, store by store vs. a twin, and the Pre-Launch Pilot Read's example is a kiosk pilot.
What if the read shows online ordering cuts into margin? Then you know where to fix it and where to stop it. A read 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, plus your app's and each delivery app's exports, 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

Every read shows where a change worked and where it didn't, so it naturally raises the next question: why? The next read usually runs on the same export you've already sent.
Reads cost less in sets of 3, and your first read will be credited toward a multi-read discount.

Scope several changes →Start with one read →How monitoring works →Foodservice read →

Sources: Seurat Analytics, "Same Brand, Different Store Experience," a national survey of 1,200 U.S. shoppers, July 2026. The online ordering figures are from the 1,087 respondents who are fast-food or c-store customers.