Payoff Reads · Deals and promotions

Did the deal make money, or just move units?

Find out what each deal did at every store, after the discount, isolated using a twin built from your stores that didn't run the deal.

0 Made money
8 No significant effect
8 Cost 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

Your report shows the deal's items sold more. It can't show whether you'd have made more money without the deal.

What you have: unit sales of the deal's items, up since the deal started, and each store's P&L.
What it can't show: what those stores would have sold without the deal. The discount goes on every unit, including the ones customers were previously buying at full price. Some of the extra units are substitution: shoppers switched to the discounted brand from a brand that wasn't on deal, often a more premium brand with a better margin. The switch suits the manufacturer running the deal, but the store gives up the better margin. Whatever the deal cost does land in each store's P&L, but a loss of $150 per week sits inside a store's normal ups and downs, so the P&L can't show it.
What the read shows: at which stores the deal made money after the discount, at which stores it lost money, and what it did to the rest of the category.

Weekly margin after the discount, vs. the twin

One store: Store 31 vs. its twin

Some losses hide inside a store's normal ups and downs.

Store 31
Normal ups and downs
Deal starts Jun 1

Sixteen deal stores, averaged vs. their twins

Average the 16 stores and the loss stands out: $150 per store per week, about $125,000 per year.

Average of 16 deal stores
Normal ups and downs
Deal starts Jun 1

See it in the 2-minute film →

A deal can lift its items and still lose money.

What lands on your desk

An energy drink deal, as Seurat reads it

Heartland Markets ran an energy drink deal at 16 of its 42 stores from June 1. Stores that didn't run the deal were used to create store twins.

The answer, in three lines:

It moved items: energy drink units rose 19% vs. the twins.
But it didn't make money: −$150 per store per week after the discount, statistically significant, or about −$7,800 per store per year.
StopRecommendation: stop it. It cost money at the 8 stores where energy drinks already sold best, and made money nowhere.

store by store (6 of 16)

Store	Energy drink units vs. twin	After the discount, per week	Verdict
Store 20, Lake Rd	+14%	−$270	Cost money
Store 4, Lincoln Way	+16%	−$240	Cost money
Store 33, Kennedy Pkwy	+17%	−$210	Cost money
Store 31, Cedar St	+21%	−$110	No significant effect
Store 9, Boyson Rd	+24%	−$60	No significant effect
Store 27, Pine St	+26%	−$20	No significant effect
All 16 stores	+19%	−$150	Cost money

By store type: at the 8 stores where energy drinks already sold best, the deal lost money. At the other 8, it had no significant effect.

by store type, per store per week

Store type	After the discount	Confidence band	Verdict
The 8 stores where energy drinks sold best	−$230		Cost money
The other 8 stores	−$70		No significant effect

Why it lost money: at the top-selling stores, most of the discounted cans were items people were previously buying at full price.

Why the table shows groups: a single store's read can miss a loss smaller than about $290 per week. Across the 16 stores, the read has more sensitivity and can detect a loss as low as about $115 per store per week. That's why we report groups as well as single stores.

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

Each deal store gets a twin: a blend of your stores that didn't run the deal, weighted so the twin matches the store's weekly sales for a year or more before the deal.
The twin shows what the store would have sold had it not run the deal. The twin went through the same weather, gas prices and local events, so the gap between store and twin isolates the effect of the deal alone.
The read checks itself. The twin has to have tracked the store closely before the deal. Placebo tests then rerun the read on a made-up start date, when nothing changed, and on stores that never ran the deal. Both should show no effect. If either shows an effect, the read isn't working properly, and we tell you.
Every store gets a result. The extra units are priced at your margin, the discount comes off, and each store gets a dollar impact, a confidence interval and a verdict.

Side note: A deal that runs at certain hours is also compared to the same stores' other hours, which catches customers who simply came in later for the deal price.

Can your chain be read?

Setup	Smallest change it can show
Half of a 30-store chain, a food category, 12 weeks after the start	About 5% of the category's sales
Half of a 30-store chain, packaged drinks, 12 weeks after the start	About 3%
One store on its own, a food category, 12 weeks after the start	About 12%
A deal at every store at once	No twin from your own stores. The read compares what the deal touched to what it didn't touch (its hours or its items), or a 6-week pause at a few stores settles it.

Rough figures from a standard formula, before we see your data. More stores, more weeks and steadier categories all lower them.

Scope this read

What we need, and what stays with you

We need:

Weekly sales by store and item, for the deal's items and the rest of their category, 1 to 2 years back
The deal's dates and discount at each store
The margin on the deal's items, as a percentage
Sales by daypart, if the deal runs at certain hours

We never ask for:

Customer or card data
Employee data
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 deals, which stores, which weeks.

Scope this read Book a call

Questions buyers ask us

What if the deal ran at every store at once? Then there's no twin from your own stores, so the read compares what the deal touched to what it didn't touch: its hours vs. other hours, its items vs. similar items. If that isn't enough, pausing the deal at a few stores for 6 weeks answers "what happens if we drop it?"
Can it read a deal a manufacturer funded? Yes. Brand-funded deals get checked for compliance, rarely for whether they added sales. The read answers that for you, and for the brand, if you choose to share it.
Can it read several deals at once? Yes, as long as they started at different stores or on different dates. Put them all in one request.
What if the read shows no lift? 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, 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 your stores measured week by week.

Scope several changes →How monitoring works →