Forms, Templates & Other Nonsense

Every consulting engagement in this industry ends with a deliverable, and the deliverable is almost always a set of documents. Manuals. Checklists. Position descriptions. Review forms. Scorecards. Onboarding packets. Standard operating procedures, bound or zipped, with your logo on the cover page.

The operator reads that stack as the value he purchased. He can hold it. He can put it on a shelf and point at it. He can hand it to a new kitchen manager and feel like he handed over something real.

I do not sell that stack. I build forms for the operators I work with, and the difference between those two sentences is the whole page.

Here is the whole thing in one line: you cannot design a system for your business that is not based on how you do business.

Read it slow, because every objection to this page dies on it. A system is the set of decisions about how the work runs. How you do business is the only thing those decisions can be drawn against. Take a system drawn against somebody else’s business and install it in yours, and you have not systematized your operation. You have installed a description of his and asked your cast to live inside it.

The Instrument The Restaurant Industry Runs On #

Templates and forms are the standard unit of exchange in restaurant consulting, and they have been for as long as the trade has existed.

The consulting side sells them because they are the only thing in the engagement that survives the invoice. A read is invisible. A conversation cannot be photographed. A document can be delivered, counted, and priced, which makes it the thing the proposal gets built around.

The vendor side sells them because they cost nothing to reproduce. Sixty-four procedures, two hundred sixty-nine pages, ninety-nine dollars, delivered the minute the card clears, editable so you can customize them. The same file goes out a thousand times, and the marginal cost after the first sale is zero.

The association and trade side gives them away as member value, because a downloadable form is the cheapest thing an organization can offer that feels like help.

The technology side embeds them. Every platform arrives preloaded with default configurations, default review criteria, default labor models, default scheduling logic. Nobody calls those templates, but that is what they are, and they are the most consequential ones in the building because they run without anybody opening a file.

The education side teaches them. A student learns the form before he ever learns the read, because the form is teachable in a semester and the read is not.

So an operator who has been in this industry twenty years has encountered the instrument from five directions, all of them reinforcing the same lesson: the document is the work.

What A Template Actually Is #

A template is a record of somebody else’s decision.

That is not a rhetorical framing. It is a description of what the artifact is. Every line on a review form is a decision about what this operation measures. Every step in a procedure is a decision about how the work runs here. Every cap in a position description is a decision about what a capable person is permitted to fix without asking. Somebody made each of those decisions, against some operation, at some point, for some reason.

The document is what is left after the deciding is done. It is what a ruling leaves behind, not the ruling itself.

Which produces the inference the whole industry runs on: this document records a working operation; therefore installing this document produces a working operation.

That inference is backwards, and everything else on this page follows from how badly.

The Reduction #

The reduction is the move that strips a decision down to its artifact and then sells the artifact as though it were the decision.

Here is the full chain, and every link matters. An operator somewhere read his own building. He saw something specific — this volume at this hour, this equipment, this cast, these Guests, this room, this menu. He made a decision against what he saw. He wrote the decision down so his people could run it. The document worked, in that building, because the decision behind it was correct for that building.

Then somebody extracted the document from the building.

What got extracted is the last step in a five-step chain. The read is gone. The operation it was read against is gone. The load it was designed to carry is gone. The reason for every choice inside it is gone. What is left is formatting.

And the formatting is the part that reproduces. The read does not reproduce. The load does not reproduce. Nothing upstream of the document reproduces, because all of it was specific to a building, and specificity is exactly what has to be stripped out to make a document work in a thousand operations.

So the reduction is: the operator wrote this document; the document ran a working operation; therefore this document runs a working operation.

The premise is true. The conclusion does not follow. And the industry has built an economy on the gap.

The Form Is The Tell, Not The Content #

Operators argue the wrong point here, every time. They defend the content. The procedure is well written. The review criteria are reasonable. A man who knows restaurants wrote it. All of that can be true and none of it is the issue.

Take any part in your operation and ask one question. Would this run identically in an operation across the country, with different Guests, a different cast, a different room, a different menu, at a different volume?

If yes, it is not carrying anything specific to your building. That is not a statement about its quality. It is a statement about its form. A part built to run anywhere was built for the intersection of everywhere, which is the thinnest version of the work — the part that stays true regardless of your building, which is another way of saying the part that knows nothing about it.

A document can be well written, legally sound, correct in the abstract, and carry no load in your operation. Those are separate properties. The industry has spent thirty years teaching operators to check the first three and never test the fourth.

The Math Inside It Was Already Yours #

Open the stack and look for the part nobody could have handed you without knowing your building.

The labor calculations are in there. There are ten of them: payroll dollars or payroll hours on top, and sales, covers, tickets, entrees out the window, or an hour of the clock on the bottom. That is the entire set. It has been printed in textbooks for forty years and every one of them is division a sixth grader does. Food cost and prime cost are the same. Your POS ran them this morning and mailed them to you.

Then count how many restaurants the rest of it can actually fit.

In 1950 the Air Force had pilots dying in accidents nobody could explain, in a cockpit that had not been resized since 1926. Gilbert Daniels measured 4,063 flying personnel on the ten dimensions that mattered for fit, and he was generous about what counted as average — the middle third of the range on each one, not an exact number. Then he counted the men who landed in the middle third on all ten at once. Zero. Not a handful. None. Cut it to any three of the ten and fewer than four in a hundred qualified.

Run that count on your restaurant. For the same stack to sell to everybody, it had to average out your seat count, your check average, your lunch-to-dinner split, how long your cast has been there, who you can hire in your town, the equipment you own, your menu, your mix, what you pay in rent and debt, and who owns the building. Be as generous as Daniels was and call the middle third average. Two out of every three buyers are already out on seat count alone, because they run ninety seats or thirty-one and not the sixty in the middle. Out again on check average. Ten times through and nobody is standing.

Which turns the pitch inside out. A thousand restaurants in forty markets over nine years gets read to you as proof it works. It is the reason it fits nobody. Every restaurant that stack had to sell to is one more thing that had to be averaged out of it, and every average moved it further from your building. The stack with the most logos on the back page is the one least able to fit the restaurant you walked out of this morning.

The Air Force read Daniels and stopped building to the average. They mandated adjustable seats, adjustable pedals, adjustable straps, so the cockpit moved to the pilot. That cost real money and they spent it anyway, because the alternative was dead pilots. Nobody selling restaurant stacks has ever run that count, so they are still shipping the 1926 cockpit, and the operator whose shoulders do not fit is told to sit up straighter.

It Got Better For A Month, And That Was You #

You remember it working. The month the stack went in, the number moved. Every operator who ever bought one remembers this, and it is the reason he bought the next one.

What moved it was you. Something got written down and looked at every week that had never been looked at every week before. You were in the building on it. Your cast knew you were on it. That produces a short-term improvement in whatever you point it at, every time, and none of it has anything to do with what was printed in the stack. A legal pad and six weeks of your attention would have produced the same month.

Then the compressor went out on a Thursday, the Sunday kitchen manager quit, you got pulled off it, and the number went back where it was. You read that as the program wearing off, so the following spring you bought a bigger one. What you were buying was somebody to make you pay attention to your own operation.

One way to settle it. Pick the number you care about, put it in front of yourself every week for six weeks, change nothing else, and buy nothing. If it moves, the stack never did the work you gave it credit for.

Six Ways The Industry Exploits The Instrument #

The document stands in for the read. The engagement produces a stack, the stack gets delivered, and nobody ever asks whether the building got read. The deliverable satisfies the obligation, so the read becomes optional, so it stops happening.

Nobody can say what any step was built to get around. Every operating part exists to work around something — a piece of equipment that could not do the job, a person who could not be trusted with a decision, a volume at an hour that made the obvious approach impossible. That reason is the load the part carries. When the part comes from inside your building, somebody remembers. When it arrives in a file, the reason was never here, and no one on your payroll can recover it. So you run workarounds for problems you do not have, and when a cast member asks why, the honest answer is that nobody knows, which teaches him something about your standards you did not mean to teach.

The same document goes to your competitor. He can buy it this afternoon. His position descriptions cap authority at the same place yours do. His variance policy escalates at the same dollar figure. His review form measures the same things. Now name what separates the two operations structurally. If both stacks came from the same catalogue, the answer is the menu and the room, and the menu is copyable in a season.

Customization is priced as the fix and functions as the cover. The document gets your logo, your product names, your dollar thresholds. That changes the language. It does not change the origin. And the need to adapt is itself the tell — a part drawn from a read of your building would not need adapting. Adapting is what you do to a part drawn somewhere else.

The document outranks the operator. This is the expensive one. An operator makes a call, says it out loud in a pre-shift, means it. Then the checklist in a cast member’s hand says the old thing, or the position description does not carry the authority his decision requires, or the system will not permit the configuration his decision assumed. Whatever is in somebody’s hand at seven-forty on a Friday is what the operation runs. The ruling took fine. It got overruled by a form.

The revert becomes proof. He does not conclude that his artifacts overruled his decision. He concludes his decision was wrong, or that his people will not execute. So he goes back to the document, and the whole episode files itself in his head as evidence that designing your own operation does not work in a real restaurant. That belief is what funds the next purchase, which is why the cycle does not break on its own.

The Sibling Move — The Fill-In-The-Blank Framework #

When an operator gets suspicious of templates, the industry sells him a framework with blanks in it.

The pitch sounds like the opposite of a template. You are not buying our answers, you are buying a structure for producing your own. Here is the canvas, the matrix, the scorecard, the four-quadrant grid. Fill it in with your specifics.

Look at what the blanks are, though. The categories are fixed. The relationships between them are fixed. What counts as a valid entry is fixed. The operator supplies nouns and the instrument supplies every decision that matters — what gets measured, what gets compared to what, what a good answer looks like, and what question is worth asking in the first place.

A framework with blanks is a template that made you do the typing. And it is a harder sell to argue with, because the operator now believes the output is his. He filled it in himself. What he did not do is decide the shape, and the shape is where the ruling lives.

The Infrastructure That Keeps Producing It #

The proposal. Deliverables get itemized and priced. A read cannot be itemized, so the proposal is built out of documents, and the engagement follows the proposal.

The purchase process. An operator comparing two consultants compares deliverable lists, because that is the only part of the two offers that looks measurable. The one who promises more documents looks like more value.

The reproduction economics. A document costs nothing to send again. Any product with zero marginal cost gets sold on volume, and volume requires that the thing work generically, which requires that everything specific be removed.

The subscription model. Ongoing access to a document library is recurring revenue. Ongoing access to somebody’s read is not a product anybody has figured out how to bill monthly.

The platform default. Every system ships with somebody’s decisions already installed, and the operator who never opens the settings has adopted all of them without a single purchase decision.

The certification. Programs teach the instrument and then credential the operator on his command of it, which means the industry’s formal training pipeline produces people whose competence is measured by how well they run somebody else’s forms.

None of that requires bad intent. It is a set of businesses doing what their economics require, and the aggregate output is an industry where the document is the deliverable and the read is a personal hobby.

This Is Not A Restaurant Problem #

Software teams buy agile ceremonies as a process template and then discover the ceremonies were designed for a team structure they do not have.

Marketing departments buy a funnel framework built against a purchase cycle that is not theirs and then optimize a stage that does not exist in their business.

Schools adopt a curriculum package built against a student population that is not the one in the room.

Hospitals install care pathways drawn against a patient mix and staffing model from a different institution and then write exception protocols to get around their own protocol.

Small businesses buy an operating manual off a franchise model and inherit constraints that only made sense when there were four hundred locations enforcing consistency across a brand.

Every one of those is the same reduction. Somebody read a situation, decided something, and wrote it down. The writing traveled. The read did not.

What The Operator Should Evaluate Instead #

Did he read my building before he proposed anything. The ones worth hiring lead with a read and you will know inside ten minutes. The ones who lead with a deliverable list are selling you what somebody else’s ruling left behind.

Does his answer depend on my Guests and my cast. If the recommendation would be the same in any operation of roughly my size, it was not drawn from mine.

What can my people do when he leaves. A capability leaves your cast able to do something they could not do before. A document leaves you holding a document.

Who is making the decisions in this engagement. Name them individually. Not who approves. Who decides.

What does this take off my schedule, my review form, my cost structure, my standard. A real engagement subtracts. If nothing comes out, nothing was decided.

How will I know it took, in my building, on a Friday, with him gone. If there is no answer to that, the engagement’s success criteria is the delivery of the stack.

What ends it. An engagement with no end is a subscription to somebody else’s judgment.

Why I Do Not Sell Forms, And Why I Build Them #

I build forms. Manuals, position descriptions, review criteria, procedures, the whole stack. Systematizing an operation is most of what I do, because an operation that runs on what three people remember does not compound and cannot hold a standard on a Friday with the operator out of the building.

What I will not do is hand you a set I brought with me.

The forms I build come from being in your building and watching what you actually do and what you actually need. Not from a read of the industry. Not from a read of operations your size. From standing on your stage at the hour that matters, in your kitchen at the volume your book actually takes, with your cast, watching where the work backs up and where it goes smooth and where somebody is carrying a decision nobody ever gave him the authority to make. That is where a form comes from. It is a record of what is already true in your operation, tightened and written down so it holds without you standing there.

That is why the order matters more than the artifact. A system has to be based on how you do business, which means somebody has to see how you do business before a single line gets written. Read first, then rule, then write it down. Write it down first and you have put somebody else’s decision in your cast’s hands and called it your standard.

Three things follow from that, and they are why the stack cannot be sold in advance.

The ruling is the one thing in your operation that cannot be transferred. I am in my 45th year of service and I still cannot rule your building from outside it. Whatever I handed you on day one would be my ruling for a building I had not yet seen, which is precisely what my work exists to get out of your operation.

The document is the cheapest part of the work and the part most likely to be mistaken for all of it. Drafting and formatting is labor — buy that from me, buy it from anybody, it does not matter who types it. What cannot be bought is the decision the draft records, and an operator admiring a stack is an operator not reading his own building.

A form installed from outside will overrule you the first time you rule against it, and you will blame yourself. I have watched that happen to competent operators for four decades. They make the right call, the artifacts in the building say something else, the operation keeps doing what the artifacts say, and the operator decides that design does not work in a real restaurant. He is wrong, but the experience felt like evidence.

Every Tool In Your Building Came From A Process, Even If Nobody Ran One #

A form is not a thing you acquire. It is the output of a sequence, and the sequence runs whether you run it or not. Buy the stack and you have skipped to the last step of somebody else’s.

The sequence is the same one every part in your operation should have come out of. Read the building. Name the load that part has to carry. Rule against it. Write the ruling down so your cast can run it without you standing there. Watch what it produces on a Friday. Re-read it against what the building is doing now. That is The Design Process, and the step that cannot be bought is the third one.

The first read is not of the building, though. It is of where you are. An operator six months from opening, an operator five years in with one room and a kitchen manager he trusts, and an operator with three rooms and a controller are not at the same place, and they cannot run the same forms. Not because one is better than another, but because a form is drawn against what the operator can decide today and what his cast can hold today.

Which is why the tools turn over as the operation matures, and turning over is the sign it is working, not the sign you got it wrong. Pre-opening you need a recipe card with a cost on it and a par sheet. At one room with a cast that has been there four years you need a review form that names what capable looks like in your building and a variance number that escalates to you. At three rooms you need a standard that holds in a building you are not in on a Friday and a way to know that it did. Each of those replaces the one before it. Run the first set at three rooms and you are back to what three people remember. Run the third set at pre-opening and you have built administration for an operation that does not exist yet.

The stack you bought has no position on any of this. It was written once, for nobody in particular, and it arrives the same whether you open in March or have been at it since 1998.

Nothing You Build Stays Built #

The second half of this is the part almost nobody tells an operator, and it is why a purchased stack fails even when the decisions inside it were once right.

Every form in your operation was drawn against a specific load. The volume at the hour it actually arrives. The cast you actually have and what they are actually capable of. The equipment you actually own. The room, the menu, the mix, the cost structure, the hours. Your ruling was correct for that load.

Then the load moves, because it always does. You add a daypart. The book gets heavier on Thursdays than it used to be. A twelve-year kitchen manager leaves and takes the part of the standard that only ever lived in his head. You take on a second room. A vendor changes a spec. Volume goes up thirty percent and the procedure that worked at the old number is now the reason tickets stack at seven-forty.

Nothing announced itself. The form still says what it said. It is just carrying a load it was never drawn against, and the gap between the two is where an operation quietly stops compounding while every document in the building still reads correct.

That is why my work does not have a delivery date where the systematizing ends. The read goes on a cadence. Every form gets re-read against the load it is carrying now, not the load it was drawn against. When you touch a part for any reason, you rebuild it rather than patch it, because a patched part is still carrying the old assumption underneath the new line.

And it is the same law running a second time. A system has to be based on how you do business, and how you do business does not hold still, so a system that never gets re-read stops being based on anything. A stack you bought cannot do any of that. It was fixed at the moment it was written, by somebody who is not coming back, against an operation that was not yours to begin with. A stack you built with me gets re-read as the operation grows, because the same read that produced it is the one that keeps it current.

You will end up with a full set of forms. It will not arrive at the start, it will not look like anybody else’s, you will be able to say what every line in it was built to carry, and it will change as your operation does.

What Changes Tomorrow #

Walk your operation and inventory every manual, every form in use, every template running the schedule, every position description, every named program, and every system on a monthly fee. Write them down. Include the things nobody wrote down, the way it has always been done that lives in three people’s heads, because those count too.

Next to each one, write the name of the person who decided it. Not who approved it. Who decided. You will end up with four piles: you decided it, somebody in your building decided it, you bought it, or nobody alive can say. For most operators the third and fourth piles together are the majority of the operation, and seeing that on one page is the first honest look they have had at their own building.

Then take the single biggest item in the bought pile and ask two questions about it. What was this built to get around, and is that thing true in my building today. If nobody can answer the first question, you have found a workaround for a problem you may not have, and it has been costing your cast time on every shift since the day it was installed.

Do not throw it out tomorrow. Put it in front of the people who run it and ask what they would do instead. That conversation will produce something no document you could buy contains, because it came from people standing in your building.

Then do the next one next week. The stack you end up with will be yours, and nobody on your street will have a copy.

Digging Deeper #

Every term in my framework lives in the Knowledge Base: kb.jeffreysummers.com

Terms used on this page: [Sameness Machine], [Default Architecture], [Designed Architecture], [Default Gravity], [Constraint Inheritance], [Configuration Arbitrage], [Vendor Capture], [Meaningfully Differentiated Value], [Ordinary By Design]

Sources Cited On This Page #

Daniels, Gilbert S. The “Average Man”? Technical Note WCRD 53-7, Wright Air Development Center, December 1952 — https://apps.dtic.mil/sti/tr/pdf/AD0010203.pdf

NASA Reference Publication 1024, Anthropometric Source Book — bibliography summary of the Daniels work — https://apps.dtic.mil/sti/tr/pdf/ADB298532.pdf