You Do Not Have A Restaurant Differentiation Strategy, You Have A Claim

Ask an operator what their restaurant differentiation strategy is and you will get one of three answers. A sentence about the food. A sentence about the people. Or a sentence about the neighborhood. All three are true. None of them are a strategy, and the operator saying them has no way of knowing that, because the sentence works. It sounds right at a chamber lunch, it reads fine on the About page, and nobody in forty-five years has ever pushed back on it in the room.

Then a place opens eleven doors down with a better sign and a bigger opening week, and the sentence stops working. Not because it was false. Because it was never doing any work in the first place.

I want to show you the specific reason that happens, and it is not the reason you have been given. The industry will tell you that you needed a clearer concept, a tighter brand, a sharper story, a niche. That answer cannot be true, and the reason it cannot be true is sitting on your street right now.

Take two operations on the same block. Same band, same price point, same year, same broadline distributor, same three specialty vendors, same payment processor, same labor pool drawing from the same few miles. Both opened within a season of each other. One of them is on its fourth year and its second location. The other one is running promotions in month fourteen. Every condition you would normally reach for to explain the gap is held constant between them. A list of conditions cannot explain a difference in results when the conditions are identical.

So the difference was produced somewhere the conditions do not reach. It was produced in the commitments — the decisions each operator made about the lease, the layout, the equipment, the hires, the contracts, the platforms, and their own hours. And the difference between those two operators is not that one of them made better decisions. It is that one of them was asking a question at each decision that the other one never asked at any of them.

Why Strategy And Claim Are Not The Same Word

Your first objection is that this is semantics. You have a differentiation strategy, I am calling it a claim, and we are arguing about labels while the covers are the covers.

Here is why a restaurant differentiation strategy and a differentiation claim are not the same object. They send you to two different places to work, and only one of those places has a lever you can still pull.

If what you have is a claim, the work is downstream. The claim is not landing, so the fix is to say it better. New photography, refreshed messaging, a repositioning exercise, a consultant who will run a brand workshop and hand you a deck. That is the entire available action set, because a claim only exists in language and language is the only place you can go to work on it.

If what you have is a strategy, the work is upstream, and it is nothing like the other list. It is a review of what you have already committed capital and capacity to, and a decision about what you commit next. That work is unglamorous, it produces no deck, and it is the only work on either list that can actually change what a Guest experiences in your building.

The two are indistinguishable while conditions are good. Covers are fine, the room has energy, the reviews are pleasant, and an operator with no differentiation at all reads exactly like an operator with a real position. They come apart under pressure and only under pressure. When the new place opens, when the trend window closes, when costs move faster than you can price, the operator with a real position has something Guests will pay to keep. The operator with a claim finds out that his Guests were never choosing him. They were choosing proximity, habit, or the absence of an alternative, and all three of those transfer to a better sign eleven doors down.

The principles follow. The mechanism underneath them is taught in full on Restaurant Physics and the arbitrage that feeds on it is prosecuted on Hacksterism, and both are linked at the bottom. What is on this page is my read, and what that read makes me do when I walk into an operation.

Sameness Is Not A Technology Problem, It Is The Entire Input Side

Most operators have now heard the argument about technology. Everybody runs one of the same handful of platforms, the platforms all do roughly the same things, and buying one does not make you different from anyone else who bought one. That argument is correct, and it is far too small.

Look at the actual list of what comes into your operation from outside it.

Your broadline distributor serves your competitors. Your specialty produce and protein vendors serve your competitors, from the same trucks, on the same days. Your POS and everything bolted to it is sold to the whole band. Your payment processing is three companies wearing different names. Your accounting and payroll platform, your insurance and benefits broker, your delivery platforms, your reservation system, your scheduling software, your marketing agency if you use one. Your labor pool is the same few miles of people who move between your building and the one down the street without changing their commute. Your leadership training came from the same small set of programs. Your consultants sell the same engagement to everyone in the band, and the advice in the trade press is published where every one of your competitors can read it for free.

Every single one of those is an input, and every single one of them arrives at your door in the same form it arrives at theirs, at a price set by volume neither of you has.

Now notice what that means, because it is sharper than the cost argument you have heard. The usual version of this is that you cannot out-buy the chains, which is true and which you already knew. The version nobody says out loud is that differentiation was never available on the input side at all. Not to you, not to the chains, not to anyone, at any price. There is no version of buying where the thing you buy makes you different, because the thing you buy is for sale.

That is not a pessimistic read. It is the most useful thing on this page, because it eliminates about eighty percent of where operators go looking.

That is why I do not open an engagement by reviewing your competitive set, and why I am not interested in your concept document. I open with the last ten commitments you made that cost real money or real capacity, and the question I am asking about each one is not whether it was a good deal.

Differentiation Only Lives In Three Places

If it is not in the inputs, it is in what you do with them. There are exactly three places, and knowing the list changes where you spend.

Composition. What you assemble out of ingredients available to everyone. Which recipe, which pairing, which sequence, which portion, which room at which hour for which occasion. The truck is the same truck. What comes off it and what you make of it is not.

The cast. The same labor pool produces wildly different output in two different buildings, and the difference is not recruiting. It is what you develop in people and what you authorize them to do in the moment without asking. A cast member who can read a table and act on it produces something no competitor’s cast produces. A cast member who has to go find a manager produces the category standard, every time, no matter how good they are.

Relational architecture. The accumulated obligation between your operation and a Guest across many visits. A competitor cannot buy it, because it is not for sale at any price, and they cannot accelerate it, because it only accrues at the rate people come back. It is the slowest of the three and the only one that compounds.

Every one of those three is built inside your building, out of decisions only you can make. None of the three is available from a vendor. Which is also why the three of them are chronically underfunded, because the decisions that build them do not arrive with a sales call attached and nobody is following up on them next Tuesday.

That is why I will not sell you a concept refresh. There is no version of that engagement where the deliverable touches any of the three places where the thing you are buying actually lives.

The Real Decision Is Allocation, And Nobody Told You That

Here is the part that gets skipped, and skipping it is why operators who genuinely try still end up at parity.

Your capacity is finite. Your money is finite, your hours are finite, and your attention is the scarcest of the three. Real differentiation requires sustained capacity on a surface over a long period, which means it lands on two or three surfaces in an operation. Not twelve. Two or three.

So your actual decision was never “how do I differentiate.” It was, and is, an allocation decision with three parts. Which two or three surfaces get real capacity. Which surfaces run deliberately ordinary. And how you keep the ordinary ones from sliding into neglect, because ordinary and neglected produce completely different outcomes and they look identical on a budget line.

That decision is unanswerable without a filter. And when an operator has no filter, the allocation still happens — it just gets made by whoever asked most recently. The vendor who called last quarter. The thing the trade press named. The competitor’s new addition that showed up in your feed at eleven at night. The operation ends up differentiating nothing and spending everywhere, and the operator experiences it as being stretched thin, which is an accurate feeling attached to the wrong cause.

That is why the first real deliverable in most of my engagements is a list of things we are going to stop trying to be good at. Operators find that harder to hear than any diagnosis I give them, and it is the single move that frees the capacity everything else depends on.

Ordinary Has To Be Decided, Or It Turns Into Neglect

The surfaces that do not get capacity still need a decision. That is the half operators skip, and it is what makes the allocation argument sound like permission to let things go.

A surface run as deliberately ordinary has a standard, a cost ceiling, and a definition of what clears. You buy the cheapest thing that meets the requirement, you do not configure it, you do not negotiate accommodations on it, and you do not spend capacity defending it. It works, it is unremarkable, and it is supposed to be unremarkable. [Ordinary By Design] is the posture, and there is nothing apologetic about it.

A surface that was never decided has none of that. It drifts, because nobody owns a standard for it. It gets worse slowly, because slowly is how unowned things get worse. And when it finally shows up as a problem, it shows up as an emergency, which is expensive at exactly the moment you have the least capacity to spend.

The difference between the two is one conversation held in advance. That is all it is.

That is why I ask, on every surface we are not funding, what standard it has to hold anyway, and who owns that standard by name. An operator who cannot answer the second half does not have an ordinary surface. He has an unowned one.

The Question You Ask On Every Surface

The filter is one question, and it gets asked before the money moves, not after.

Does this let my operation produce value a Guest can feel and name, that the operation down the street cannot produce off the same inputs?

Yes means this is a surface worth real capacity, and it gets spent deliberately. No means the surface is table stakes and gets bought as table stakes. Both answers are useful. The only unusable answer is the one you never got because you never asked.

Every surface has its own version of that question and its own tell.

The tool. Not whether it works or what it costs. Whether it lets you produce something a Guest can feel that your competitor cannot produce off the same platform. When the honest answer is no, the tool is table stakes and should be bought like table stakes — cheapest thing that clears, no configuration, no accommodation. What operators do instead is negotiate hard for custom setup on a platform that was never going to differentiate anything, then pay for that setup twice, once in fees and once in the complexity it adds to every shift after.

The person. Not whether they can do the job. Whether they can produce something in front of a Guest that nobody else on your cast produces. Almost all hiring asks the first question. The second one is the only half that touches differentiation, and it is also the half that should decide who you develop and who you authorize.

The contract. Every contract you sign either widens what your operation can do or narrows it. Operators sign narrowing contracts constantly, and not out of carelessness. The narrowing arrives later, as a constraint you had forgotten was chosen. The savings arrive now, as a number on a page. That asymmetry is the whole reason it keeps happening.

The lease. This is the most permanent differentiation decision you will ever make, and it is the one most often settled on rent per square foot and nothing else. The lease decides which occasions your room is physically able to host. That is the ceiling on everything downstream of it, set before a single Guest walks in, and no amount of good work inside the room raises it.

The building. The layout either produces the GX every shift or fights it every shift, for the length of the lease. Decide it on buildout cost alone and you inherit the fight at full price and pay it in labor, in timing, and in the specific moments where a Guest can feel your operation straining.

The stack. Same logic as the tool, at the level of the whole assembly. Every addition brings its own reporting, its own login, its own integration, and its own claim on the attention of the people who are supposed to be producing hospitality on the stage.

The equipment package. Specified on capacity and price, almost never on what it lets the menu do that a competitor’s package cannot. The equipment decision quietly sets the boundary of your composition for a decade, which means it is a differentiation decision whether or not it was treated as one.

The tell is identical across all seven. The answer was never yes and it was never no, because the question was never put. What got asked instead was “can we afford this” or “do we need this.” Both are legitimate questions that have to be answered eventually. Neither one is the differentiation question, and an operator who substitutes one for the other on every surface will arrive, entirely rationally and with a clean conscience, at an operation assembled from the same components as everyone else and priced as though it were not.

That is why I ask to see your lease before your P&L. The P&L tells me what happened. The lease tells me what was ever possible.

It Has To Land On All Five, Not Just The One You Like Talking About

A difference that works for the Guest and nowhere else is not a strategy yet. It has to land across every actor your operation touches, and there are five of them: you, the Guest, the cast, your vendors, and the community.

Walk your claimed difference through all five and say out loud what each one gets from it.

If it lands for the Guest but not the cast, they will not produce it consistently, and you are one resignation away from losing it entirely. You will read the inconsistency as an accountability problem. It is not one.

If it lands for the cast but not the numbers, you have a good idea still waiting on the rest of the work, and it will be cut in the first tight period.

If it lands for the Guest and the cast but your vendors cannot support it — wrong delivery days, wrong minimums, wrong lead times — it will fail on the specific shifts where it matters most.

If it has nothing in it for the community you sit inside, it will not travel by word of mouth, which is the only distribution an independent gets for free.

And if it lands for the other four but you do not personally believe in it, you will be the one who cuts it, the first time something else needs the money.

That is why I do not accept a differentiation answer that only has a Guest in it. Five actors, five answers, said out loud, before we spend anything.

What This Costs Me To Say

One more thing, and it indicts my own industry, including me.

Almost everything sold to you as differentiation is sold on the input side, and it is sold there because that is where the sales calls are. The platform vendor, the agency, the broadline representative, the equipment dealer, the consultant with a positioning workshop. Every one of them is offering you something that is, by definition, for sale to your competitors. That is not a conspiracy. It is just what happens when the entire supply side of an industry makes its living on inputs and the only real differentiation available lives in decisions nobody can invoice.

So here is what I will not do. I will not tell you which POS to buy, because I do not think that decision is capable of differentiating you and I am not going to take money to pretend otherwise. And I will not take an engagement whose deliverable is a positioning statement. A sentence about who you are is not a strategy. It is the thing you get to write after the commitments underneath it have already made it true.

That is why I consult with operators rather than with restaurants. The restaurant cannot ask the question. Only the person signing the commitments can.

Your Restaurant Differentiation Strategy Is A Number, And You Have Never Counted It

Seven tests. Run them in order, on your own operation, with nobody watching.

The substitution test. Write your differentiation claim down in one sentence. Now swap in the name of the operation across the street and read it again. If the sentence survives the swap intact, it is not a strategy and never was. A real claim is one a competitor would refuse to make, not one every competitor already makes.

The consideration set test. Pick one occasion — a Friday two-top, a Tuesday family table, a weekend group of eight. Name the three other places a Guest seriously considered for that specific occasion, including staying home. Then say what you produced that those three could not. An operator who cannot name the three is comparing himself to last year, which is the comfortable comparison and the one that decides nothing.

The same-inputs test. Take your claimed difference and ask whether the place down the street could produce it using your broadline, your platforms, your equipment, and your labor pool. If the answer is yes and the only thing stopping them is that they have not thought of it, your difference has a shelf life measured in how long it takes them to think of it.

The naming test. Ask a returning Guest what makes your operation different. No prompting, no list to choose from. What they say is the differentiation that exists. What you hoped they would say is the claim. The gap between the two is the exact size of the work in front of you. And if the answer comes back as a category word — good food, nice people, always solid — you are at parity and the Guest is being polite about it.

The decision ledger test. List the last ten commitments you made that cost real money or real capacity. Against each, name which question was actually asked at the moment you committed. Three possible answers and no fourth: the differentiation question was asked and cleared, the surface was bought deliberately as table stakes, or the question was never put.

The substitution-of-questions test. For every commitment where the question was never put, name the question you asked instead. It will almost always be “can we afford this” or “do we need this.” The test is not whether you asked those. The test is whether you asked them instead of the differentiation question or after it. The order is the entire diagnostic.

The migration test. Name the thing that made you different two years ago and is now standard in your band. Then name what has replaced it. An operator who can answer the first and not the second is holding a position the market has already repriced without telling him, and that is the entry condition for [Static Decline].

How the score sorts is simple. Three or fewer of your last ten commitments with a real answer means you do not have a differentiation problem in the marketing sense, you have never run the filter. Half means you are running it on the decisions that feel strategic and skipping it on the ones that feel operational, which is where the lease and the equipment package live. Eight or more and the filter is running — at which point the conversation becomes whether the two or three surfaces you funded are the right two or three, which is a much better problem and a much shorter engagement.

What You Do Monday Morning

One sheet of paper, one column, no software.

Write down the last ten commitments your operation made that cost real money or real capacity. Not a year’s worth — the last ten, whatever they happened to be. A platform renewal. A hire. An equipment purchase. An agency retainer. A menu rebuild. A buildout. A standing block of your own hours.

Next to each one, write a single word: cleared, ordinary, or never. Cleared means you asked whether it would let you produce something a Guest can feel and name that the place down the street cannot, and the answer was yes. Ordinary means you decided on purpose that this surface was table stakes and bought accordingly. Never means the question was not asked, and you will know which ones those are because you will remember what you asked instead.

Count the nevers. That number is your restaurant differentiation strategy as it actually exists today, and for most operators it is the first time they have seen it as a number instead of a feeling.

Then take the largest never on the list — the one that cost the most money or the most capacity — and ask the question late. If the answer is no, that surface is table stakes you bought at differentiation prices, and the move is to stop spending capacity defending it and reprice it as what it is when it comes up for renewal. If the answer is yes, then you have live differentiation you arrived at by accident, which means nothing in your operation is currently protecting it. Name it, fund it, and make sure your cast knows it is the thing.

Either result is worth the twenty minutes. A no frees capacity. A yes tells you what to defend.

The Market Prices The Claim, It Has No Opinion On The Strategy

You do not have a differentiation problem. You have a question you were never told to ask, which you have now skipped at somewhere between six and ten of your last ten commitments, and the parity you are looking at is the sum of those skips.

That is the difference between a strategy and a claim, and it is not a word game. A claim is a sentence you write once and then defend. A strategy is a question you ask every time money or capacity moves, which means you can start running it on the very next decision in front of you, with no budget, no vendor, and nobody’s permission.

The market is going to price your claim for you eventually. It has no opinion at all about your strategy, because a strategy is not a thing it can see. It only ever sees what the strategy produced.

Digging Deeper

Every term in my framework is defined in full in the Knowledge Base: https://kb.jeffreysummers.com/

Terms used in this piece: [Meaningfully Differentiated Value], [Ordinary By Design], [Static Decline], [Values Of Sameness], [Safest Mediocre Execution], [Differentiation Economics], [Structural Scale], [Competitive Value Read], [Positioning Capital], [Unique Experience Proposition], [Authority To Execute], [Connection Floor], [Constraint Inheritance], [Configuration Arbitrage], [Vendor Capture], [Contract Constraint], [Two Roads]

The architecture beneath this read, taught in full: https://physics.jeffreysummers.com/

The arbitrage this pattern feeds, prosecuted in the wild: https://hacksterism.jeffreysummers.com/