Nothing On Your P&L Is Structural

Nothing On Your P&L Is Structural

Every line on the P&L was set by a decision. Here’s how to read the ones that grew while nobody decided.

Nothing On Your P&L Is Structural

Nothing on a restaurant P&L is structural, because every line on it was set by a decision, and the only thing a line can tell you is whether that decision was made by design or by default.

A hotel P&L session at a lodging conference this fall named the costs climbing fastest: booking fees, card commissions, and technology subscriptions that are never capitalized, so every increase lands straight on the P&L. Then it called them structural, costs that “track guest behavior and technology choices” and are “unlikely to ease on their own.” The same lines run through every restaurant P&L as third-party commissions, processing fees and a [Vendor Stack] billed by the month. They look structural from inside the operation because nobody designed them. Each was added one reasonable decision at a time, and none of them ever came back up to be decided again. This piece is the architecture for both sides of that P&L: the costs that follow the operator’s defaults, the revenue that follows the Guest’s reasons, and how to design each so the operator owns the door, the relationship and the margin instead of renting them.

The Session’s Own Words Give It Away

The numbers in the session are real. Total revenue per available room grew 6.2%, enough to offset high labor cost growth. Conference and events, a large and highly profitable segment, grew and drove margin. Guests spent more in house. The warning was right too: RevPAR alone hides the pressure, because a strong top line can mask cost lines rising just as fast.

The read stopped one word short. A cost that tracks a technology choice tracks a choice, and a choice can be made again. A booking fee rises because more Guests come in through someone else’s door, and that behavior was trained by the operation that let the first contact move there. Calling those lines structural doesn’t describe them. It’s how a default protects itself from being decided.

Why A Default Looks Structural

The operator who calls a line structural isn’t careless. He reviews his P&L every month. The problem is where the review runs from. My framework calls it [Default Perspective]: the examination runs inside the frame that produced what it’s examining, so every finding confirms the installation. The commission line arrives, the review checks it, and it passes. It was there last month. It’s there this month. Every operator in the market has one. Same rotation, same finding, same confirmation.

Leaving that loop takes one question the default never asks of any line: who decided this, and when was it last decided again? A line that can’t answer the second half is a default, even if it started as a design. Design is never set. It moves with the restaurant’s growth, and a line decided once and never re-read has stopped being designed.

The read also comes before the channel. What the restaurant is for, and who its ideal Guest is, decide which doors it needs and what each one is worth to him. An operation that skips that read lets the channels decide for it, and the commission line is the receipt.

The Door Is Part Of The Product

The Guest’s first contact is part of the product. When a third party holds it, the first impression, the login, the record and the next visit all sit on someone else’s door. The restaurant’s name is on the order, its reputation brought the Guest, and the margin goes to whoever owns the door.

Every input that moves margin has an owner, and exposure and control belong to the same party. The commission rate is owned by the platform. The operator carries the exposure and holds none of the control. On his own door, the Guest’s total is the restaurant’s price, there is no fee line to judge, and the relationship stays where it was built. A cheaper rented door is still a rented door. Cheaper rent is still rent.

Processing works the same way. The fee is charged on the whole ticket, tip included, so the restaurant pays to process money that was never its revenue. That line is a contract somebody signed, with terms somebody can read and decide again. On most P&Ls nobody has looked at it since the day it was signed.

The revenue side is designed the same way, or it isn’t designed at all. Private dining and catering are the restaurant version of the session’s conference and events line, and they do the same double duty: a growing line and the most profitable one. The reason is architecture. They are booked ahead, prepped to a count and staffed to a known number. It’s the one revenue stream in the building where demand, purchasing and labor are known before the day starts. That isn’t revenue the restaurant captured. It’s revenue the restaurant designed.

The session’s top takeaway was capturing more of what Guests spend in house. Capture is the wrong architecture. The Guest spends more in house when the experience gives him a reason to: the bottle that fits the table, the dessert he didn’t plan on, the private room for the dinner that matters. That spend is Guest-granted value. He grants it when the experience earns it, and he takes it back the night it doesn’t.

The Cast Is The Door Nobody Can Rent

The reason has to come from someone. The cast talks about the value: what the dish is, why the bottle works, what the room is for. That isn’t an upsell script. A script pushes an item at every table. A cast member who knows the product and reads the Guest explains what something is worth to the Guest in front of him, and the Guest decides.

The cast is also the one part of the operation a rented door can’t copy. A platform can carry the menu, the photos and the order history. It can’t carry the server who knows the regular’s name and his order, or the host who remembers why he booked last time. The Guest who is known comes back through the door where he is known. The relationship stays with the operator only as long as the cast holds it.

Read The Seat Hour, Then Read The Stack

RevPAR has a restaurant equivalent, and it’s sharper. RevPASH is revenue per available seat hour: total revenue divided by available seats times hours open. Eighty seats, a five-hour dinner service and $6,400 in revenue come to $16 per available seat hour. That number doesn’t care how busy the night felt. It measures what the space produced against what it could have produced, and the gap between the two is the Lost Opportunity Tax, running every hour the doors are open whether anyone measures it or not.

The hour is the part that matters. A seat at 7 p.m. Saturday and the same seat at 2 p.m. Tuesday are two different inventories. Read RevPASH by daypart and the demand cycles show themselves: where the seat hours fill, where they sit empty, and where the lowest demand cycle is. That cycle is where the base price belongs. RevPASH can also rise on price alone while the same seat hours sit empty. The top line climbs, the Guest count doesn’t, and the cost lines rise underneath both. That’s the session’s warning about RevPAR, and it’s truer in a restaurant.

Then read the stack. A [Vendor Stack] is the operator’s complete inventory of the tool layer under every function: POS, reservations, scheduling, inventory, payroll, loyalty, reviews, marketing, online ordering. Each layer has a tool, a labor model, a cost across its full life and an exit cost. Subscriptions are never capitalized, so each one lands on the P&L as its own small line, and nobody adds them up. Total them into one number and the stack becomes a decision again.

Every layer then answers for itself: what does it deliver to the Guest’s experience, and would you buy it again today at today’s price? Technology amplifies the operation it’s installed in. On a designed operation it compounds the design. On a default operation it runs the default faster. And because every operator buys the same stack from the same vendors, whatever efficiency it creates is available to everyone and competes away. That’s the [Sameness Machine]. A tool earns its line only when it serves a design the competitor down the street doesn’t have.

The Diagnostic

The Decision Test. List every cost line that grew faster than sales this year. Can you name who decided each one and when it was last decided again? Yes for every line, or no.

The Door Test. Do you know what share of this month’s Guests made their first contact through a door you own? Yes, or no.

The Stack Total. Can you state the monthly total of every subscription in the building, and what each one delivers to the Guest? Yes, or no.

The Seat Hour Test. Do you know RevPASH by daypart, and which daypart is your lowest demand cycle? Yes, or no.

The Reason Test. When check average rose, did the Guest get a reason or a script? If you can’t say which, it was the script.

Every no is a line running by default. Every yes is a line somebody designed.

Profit Keeps The Score

The default P&L pays rent on a door it could own, fees on money that was never its revenue, and subscriptions nobody has totaled, and it calls all of it structural because the frame that reviews the P&L is the frame that produced it. The designed P&L reads every line as a decision, keeps the first contact on its own door, earns the spend the Guest grants, and puts its seat hours to work at the right price for the cycle. The difference between the two is the Lost Opportunity Tax, and it never shows up as a line of its own. It shows up as the margin the default never had. Nothing on your P&L is structural. Every line is a decision, and the only question is whether you’re the one who made it.

If you like this move, you should subscribe to The Monday Morning Move — one move, one asset, one test, every Monday: https://www.fromtheplaybook.com/subscribe

Digging Deeper

  1. Your Door Is The Only One That Keeps The Guest: The Olo App, DoorDash And Uber Eats Want Him On Theirs — https://physics.jeffreysummers.com/your-door-is-the-only-one-that-keeps-the-guest-the-olo-app-doordash-and-uber-eats-want-him-on-theirs/
  2. Margin Is Engineered — https://physics.jeffreysummers.com/margin-is-engineered/
  3. Every Input That Moves Your Margin Has An Owner — https://physics.jeffreysummers.com/every-input-that-moves-your-margin-has-an-owner/
  4. The Tool Stack Is Not A Framework — https://hacksterism.jeffreysummers.com/the-tool-stack-is-not-a-framework/
  5. The Lost Opportunity Tax — https://hacksterism.jeffreysummers.com/the-lost-opportunity-tax/
  6. Catering: The Growth Engine Hiding In Plain Sight — https://kb.jeffreysummers.com/docs/5-x-catering-the-growth-engine-hiding-in-plain-sight/
  7. By Design Or By Default — https://kb.jeffreysummers.com/docs/by-design-or-by-default/
  8. Structural Only Because Nobody Designed Them (LinkedIn) — https://lnkd.in/p/gV54DWzT

Terms used in this piece: By Design Or By Default, Vendor Stack, Default Perspective, Lost Opportunity Tax, Sameness Machine. Definitions in the Knowledge Base.


Discover more from Restaurant Physics

Subscribe to get the latest posts sent to your email.

Jeffrey Summers
Jeffrey Summers
physics.jeffreysummers.com

I've been in the restaurant business for 45 years. My work has been to design better restaurants by developing better operators. Smarter thinking wins. Period. The operator who puts it to work runs a better business. I have written several books on it, The Operator's Playbook among them.

Related Posts
Leave a Reply