Every month the industry publishes a performance read, and every month operators read it as a weather report about someone else. This past June the read said comparable sales rose 1.8% while traffic fell 1.9%, and it described the result as resilient sales driven by sustained higher average check sizes. Traffic had already fallen 2.0% in May and 3.3% the previous December. By one count, operators reported a net traffic decline in fifteen of the previous sixteen months.
Read that sentence again with the adjective removed. Sales rose because the people who still came spent more. Fewer of them came. The word resilient is carrying the entire argument, and it is carrying it on a metric that cannot see the thing that got worse.
The Instrument Produces The Counsel
This is [Measurement Asymmetry] operating at industry scale. Per-person average, average check, ticket average — every one of them is a ratio whose denominator is the occasions the operation won. A lost occasion does not enter that ratio as a zero. It is absent from the data entirely. Absence reads as neutral. So an operation can shed a quarter of its Guests, hold the wallet of the remainder, and print a number that goes up and to the right on every dashboard it owns.
Run the full roster of metrics the industry actually manages to. Per-person average. Average check. Covers. Turns. RevPASH. Sales per square foot. Sales per labor hour. Every denominator in that list is capacity or headcount. Not one of them has a Guest in it. The instrument set measures how thoroughly the operation harvested whoever walked in. It is structurally silent on everyone who decided not to.
That silence is not a data-quality problem to be fixed with better reporting. It is the shape of the instrument. And because the counsel the industry produces is generated downstream of the instrument, the counsel inherits the blind spot. This is [The Dashboard Trap] with a trade-press distribution channel attached.
Forty-Four Years Of The Same Substitution
I have been saying forever and a day that this industry does not have a demand problem it can price its way out of. It has a unit problem. It measures dollars per attendance and then makes growth decisions with a number that only knows about attendance it already got.
I am not going to pretend this is new. What is new is the confidence. We have arrived at a point where a shrinking base is reported as resilience, and nobody in the room asks resilient against what. The number went up. The room moved on.
Four Readings, Each One Wrong At The Mechanism
Below are the four readings the current data is producing across the industry. Each one is individually defensible, each one compounds with the others, and the stack of them produces an operator who is optimizing his way out of business with a straight face.
One — Resilient Sales
What the reading says. Sales are up. Traffic is soft but manageable. The consumer is under pressure and the operation is holding.
Why it fails. Sales up on traffic down is not a mixed result, it is a specific result: the operation is extracting more from a shrinking base. Both halves of that sentence are load-bearing and only one of them is in the headline. At 1.8% sales growth against 1.9% traffic decline, the entire gain is spend per remaining occasion. There is no volume in it. And spend per occasion has a ceiling that arrives long before the base stops shrinking.
The cumulative impact. Because the sales line is green, the operation reads its condition as stable and defers the structural work. Every month the base narrows and every month the read comes back green, so the deferral renews itself. That is [No Static Achievement] running in reverse — a position eroding underneath a number that says it is being held.
The move that would work. Stop reporting sales and traffic as two lines on the same page and start reporting them as one number: occasions won, and dollars per occasion, in that order. Growth is the first number. The second is a performance read on the first. When the first falls and the second rises, the operation names it out loud as a smaller operation with a better harvest, because that is what it is.
Two — The Consumer Is Value-Seeking
What the reading says. Guests are trading down. The response is price: discounts, value menus, bundles, limited-time offers built to hit a number.
Why it fails. The substitute channel is not winning on price. Thirty-seven percent of shoppers say grocery deli-prepared food is just as affordable as restaurant options — as affordable, not cheaper. And when researchers asked why people buy it, the drivers came back functional rather than purely economic, with convenience and time savings cited most. The channel is winning at parity price on a variable that is not price. So the discount is being fired at a variable that was never in contention. That is the [Transactional Affordability Lie] with outside data attached to it: the operator concludes he is expensive because the alternative is chosen, when the alternative is chosen at the same money for a reason he never measured.
The cumulative impact. The discount lands on the base that already stayed. Those Guests were not the ones deciding against the operation, so the discount buys nothing incremental and gives away margin on volume that was already committed. The traffic line does not move, the check line finally does move — down — and now the one number that was carrying the story stops carrying it.
The move that would work. Before any price action, run the substitute list. Name the three specific alternatives that captured the occasions the operation used to hold, and name the variable each one won on. If the variable is convenience, timing, or the ability to finish the meal at home, no price move addresses it. Build against the variable in contention, not the one that is easy to change in the POS.
Three — Bigger Baskets
What the reading says. Order sizes are up. Guests are spending more per visit. That is an upsell win.
Why it fails. Check averages rose 3.5% in May, the strongest in over a year, and the composition of that growth is the whole story: quantity per transaction was up 2.2% against price up 1.2%. The growth is bigger orders, not higher prices. Bigger orders on fewer visits is not an upsell. It is occasion consolidation — the household that used to buy three separate occasions now buys two, and stacks what it needs into a larger order. The trade named the pattern honestly, bigger baskets, fewer visits, and then reported the first half as a strategy.
The cumulative impact. The operation now has an incentive structure pointed at the wrong outcome. Cast members are coached to build bigger tickets, managers are measured on average check, and the whole apparatus gets better at harvesting the consolidated occasion — which accelerates the consolidation. Three occasions become two, then two become one, and every step of the way the check average confirms that things are improving.
The move that would work. Split the check-average read into its two components every period: how much of the movement is price and how much is quantity. Quantity growth on flat or falling occasion counts is a consolidation warning, not a performance win, and it gets reported that way to the cast and to the operator in the same breath.
Four — The Competition Is The Restaurant Down The Street
What the reading says. The comp set is other restaurants. Occasions lost went to a competitor with a similar concept in a similar radius.
Why it fails. The grocery deli re-aimed. Between 2017 and 2025, the share of prepared-food purchases substituting for a restaurant meal went from 12% to 28%, while substitution for a home-cooked meal fell from 50% to 30%. That is a category that stopped competing with the home kitchen and turned to face the restaurant. It is running restaurant menus to do it — unit growth in the deli case is concentrated in prepared meats, appetizers, salads, pizza, sushi, and fully cooked meat. Twenty-three percent of shoppers say they are stopping at fast-food and fast-casual less often, while 85% purchase the same amount of deli-prepared or more.
And the sharpest part of it is the part no restaurant metric can register. Fifty-three percent of buyers build a hybrid meal — a prepared entrée plus something finished at home. That occasion was not lost to a competitor and it was not kept. It was split. Half of it left. There is no line on any restaurant report where a half-occasion appears.
The cumulative impact. The operator benchmarks against a comp set that is losing the same occasions he is, sees himself performing in line with peers, and concludes the softness is macro. Peer-relative performance is the most expensive comfort in this business. An entire comp set can decline in formation and every member of it can pass its own benchmark.
The move that would work. Rebuild the comp set around the occasion instead of the concept. For a given daypart, list every channel that could have taken the eating occasion — the deli case, the c-store, the fridge, the desk drawer, the canned drink. Deli purchases concentrate between noon and five in the afternoon, and about 60% of people decide lunch the same day. That is a decision window the operation can be present in or absent from, and right now most operations are not competing in it because they do not believe they are in it.
What Is Actually Producing The Counsel
Four wrong readings from four different corners is not four mistakes. It is one instrument.
Every reading above is what you get when the only numbers in the room have capacity or headcount in the denominator. The instrument can describe the harvest with great precision and it cannot see the field. So the counsel it generates is always about improving the harvest — better upsell, tighter pricing, sharper mix, higher throughput — and never about the field, because the field is not in the data.
That is why the counsel feels sound. It is internally consistent. It is rigorous against the numbers it has. And it is aimed at the wrong problem, which is the only failure mode a rigorous process cannot detect in itself.
One more piece of discipline, and it cuts against my own argument, so it belongs here rather than in a footnote. The most useful body of research on the grocery channel is published by the grocers’ trade association. It is credible and I have used it above, but across six sources on that category, not one named a single weakness of grocery prepared foods. Evidence that only points one direction is telling you something about who assembled it. Cite it, use it, and name who owns the instrument — including when the instrument favors your case.
The Diagnostic
Five tests. Run them against your own last four periods, tonight, off numbers you already have.
Test One — The Occasion Count. Pull occasions won per period for the last twelve periods, not sales. Guest counts, covers, transactions, whatever your system calls the unit. Is the line rising or falling. If it is falling while sales rise, every growth conversation in the building has been running on the wrong number.
Test Two — The Composition Split. Take your check-average movement over the same twelve periods and split it into price and quantity. If quantity is doing most of the work while occasion counts fall, you are being paid for consolidation and reading it as performance.
Test Three — The Substitute List. Name the three specific alternatives that took the occasions you lost. Not categories — specific. The deli case at the store two miles out. The breakfast program at the c-store on the commute. The container in the Guest’s own refrigerator. Then name the variable each one won on. If you cannot fill this out, you do not know who you are competing with.
Test Four — The Half-Occasion Read. For your top three dayparts, ask whether the occasion you lost was lost whole or split. A Guest who buys an entrée elsewhere and finishes the meal at home was in contention and you were partially in the running. A whole-occasion loss and a half-occasion loss require different moves, and neither one appears in your data.
Test Five — The Denominator Audit. List every metric your managers are measured on. Write the denominator next to each one. Count how many have a Guest in them. That count is your answer.
How the score reads. Zero denominators with a Guest in them is the industry default, which means the score is not a verdict on you, it is a verdict on the instrument you inherited. One or two means somebody in your operation has already noticed. Three or more and you are running a read most of this industry does not have.
What You Do Monday Morning
Pick your weakest daypart. Pull the occasion count for that daypart for the last twelve periods and put it on one page by itself, with no dollar figure anywhere on the page. Then bring the person who runs that daypart in and ask one question: where did those occasions go.
You will not get a complete answer. You will get the beginning of one, and it will be the first time that conversation has happened in your building with the dollars removed from the room. That is the whole point of the exercise. The dollars are what let everyone agree that things are fine.
The Closer
Sales up on traffic down is not resilience. It is a smaller operation being paid more per visit, on its way to being a smaller operation being paid the same per visit, and the metric that reported the first half of that sentence will report the second half as a surprise. Your sales are up because fewer people came. Both clauses are true. Only one of them is a strategy.
To understand the ideal state, go to Restaurant Physics.
Digging Deeper
Positions on the record
-
Why Too Expensive Is Never About The Price — https://hacksterism.jeffreysummers.com/why-too-expensive-is-never-about-the-price/
-
Loyalty Is Not A Discount Problem — https://hacksterism.jeffreysummers.com/loyalty-is-not-a-discount-problem-2/
-
Shrinkflation Is A Guest Contract Violation, Not A Brand-Preservation Strategy — https://hacksterism.jeffreysummers.com/shrinkflation-is-a-guest-contract-violation-not-a-brand-preservation-strategy/
-
The [VoC] Industry Sells You The Silence Of Your Guests — https://hacksterism.jeffreysummers.com/the-voc-industry-sells-you-the-silence-of-your-guests/
-
There Is No Such Thing As Guest Experience Preference — https://hacksterism.jeffreysummers.com/there-is-no-such-thing-as-guest-experience-preference
-
Administered Pricing Without A Pricing Department — https://physics.jeffreysummers.com/administered-pricing-without-a-pricing-department-what-independent-operators-actually-face/
-
QFO: What Else Can I Do But Discount — https://physics.jeffreysummers.com/qfo-what-else-can-i-do-but-discount/
-
The Occasion Is The Unit — https://physics.jeffreysummers.com/the-occasion-is-the-unit
-
The Operator’s Contract — https://jeffreysummers.com/the-operators-contract/
Term definitions from the Knowledge Base
-
[Share Of Stomach] — https://kb.jeffreysummers.com/docs/share-of-stomach/
-
[Share Of Experience] — https://kb.jeffreysummers.com/docs/share-of-experience/
-
[Measurement Asymmetry] — https://kb.jeffreysummers.com/docs/measurement-asymmetry/
-
[The Dashboard Trap] — https://kb.jeffreysummers.com/docs/the-dashboard-trap/
-
[Transactional Affordability Lie] — https://kb.jeffreysummers.com/docs/transactional-affordability-lie/
-
[Transactional Pricing Substrate] — https://kb.jeffreysummers.com/docs/transactional-pricing-substrate/
-
[The Transactional Substitution Kit] — https://kb.jeffreysummers.com/docs/the-transactional-substitution-kit/
-
[Lost Opportunity Tax] — https://kb.jeffreysummers.com/docs/lost-opportunity-tax/
-
[No Static Achievement] — https://kb.jeffreysummers.com/docs/no-static-achievement/
-
[Hacksterism] — https://kb.jeffreysummers.com/docs/hacksterism/
-
[The Read] — https://kb.jeffreysummers.com/docs/the-read/
-
[By Design Or By Default] — https://kb.jeffreysummers.com/docs/by-design-or-by-default/
