AI assistant · N5 Complete system
Contrarian Hook Generator for Restaurants
Contradicting a belief everyone repeats opens the video; backing it with a figure from your own kitchen is what keeps the argument from turning against you. This one delivers the belief map with its risk rating, the counter-claim for each belief with house proof attached, and the defence sheet for the comments that arrive afterwards.
N5Complete system
Several chained pieces you install and operate over time, with their review cadence. It changes how the business runs, not one decision.
◆ This assistant 110 of the 164 published assistants are at this levelIs this for you?
What it does and what you get
- Every owner knows the three things people say every week that strike him as false, and he keeps them to himself so as not to look like a picker of fights.
- Fair enough: most restaurant controversies collapse because nobody backed them with anything.
- This piece turns that irritation into material under one condition that orders the whole job: every contrarian claim must stand on a figure from your own restaurant or on the way it operates, and whatever cannot stand is not softened, it is dropped.
- On top comes the part almost nobody writes before posting: how much reputational risk is being bought, what gets answered in the comments, and what never does.
What you end up with
The belief map with who holds each belief, how widespread it is, its reputational risk and the decision to run, wait or drop it; the counter matrix with house proof beside each row, the dish behind it, the network where it gets said and the 14-day metric; and one defence sheet per hook, with answers pre-written for the predictable comments and the limit of what never gets answered.
EXAMPLE This is what you getAn example built with Sazón de Origen, the house's fictional restaurant.Not your data: your result is built with YOUR restaurant's.See the full example →Sample result
The owner asked this: «Sazón de Origen, contemporary Colombian food in Medellín: 2 locations, 6 years, 24 people, average check of 48.000 COP and 34% food cost. I sell in the dining room, through my own delivery and one delivery platform, and I post on Instagram and WhatsApp. I am tired of being told that home-style food has to be cheap and that delivery arrives exactly like the plate in the dining room. I want hooks that say the opposite and that I can defend with what I actually do in my kitchen, without picking a fight with anyone.»
| Belief, as people say it | Who holds it | How widespread | Reputational risk | Decision |
|---|---|---|---|---|
| «Home-style food has to be cheap» | Weekday lunch guest | Very widespread | Medium | Runs: the recipe costing backs it |
| «Delivery arrives just like the dining room plate» | Own-delivery guest | Widespread | Low | Runs: the delivery menu is already trimmed |
| «You have to be on every delivery platform» | Owners in the trade | Very widespread among peers | Medium | Runs: the commission comes out of this till |
| «The bigger the menu, the better the restaurant» | Weekend guest | Very widespread | Low | Runs: the menu went from 28 to 19 dishes |
| «A restaurant that is empty at lunch is a bad restaurant» | Trade and guests | Medium | High | Waits: Tuesday does 45 covers and the counter would read as an excuse |
How it was decided. All five came from the table, the peer group and the last 90 days of reviews; none was invented to have something to say. The first four run because a house figure holds them up. The fifth waits, not because it is false, but because anyone hearing it will look at the Tuesday dining room and find the evidence against it. It is filed with its condition written down: it returns when Tuesday holds 60 covers for 4 straight weeks.
| No. | Belief being contradicted | The counter, first line | The house proof shown on camera | Dish or figure | Channel | 14-day metric |
|---|---|---|---|---|---|---|
| 01 | Home-style food has to be cheap | «Home-style food is not cheap food: it is the food that spends 6 hours in the pot.» | Beans go on at 5:00 a.m. and the meat is weighed: 220 grams per platter | Bandeja de la casa, 34.000 COP | Platters sold Tuesday to Friday (target: from 62 to 80 a week) | |
| 02 | Delivery arrives like the dining room | «There are three dishes we will not send you by delivery, and it is not laziness.» | Trucha al ajillo loses its point after 25 minutes in transit; it was pulled from the delivery menu in writing | Trucha al ajillo | Instagram and WhatsApp | Cold-plate complaints (target: from 5 to 2 in 90 days) |
| 03 | You have to be on every platform | «We pulled dishes off the platform on purpose: the commission was eating the plate.» | At 34% food cost, on the platform the trout left less contribution margin than the same trout on own delivery | Trucha al ajillo | Own-delivery orders (target: +15% in 4 weeks) | |
| 04 | The bigger the menu, the better | «We cut 9 dishes off the menu and the sancocho got better.» | The menu went from 28 to 19 dishes in 2024; protein waste dropped and the pass stopped fighting 6 preparations at once | Sancocho de tres carnes, 42.000 COP | Sancochos sold on Sunday (target: 55) |
What holds the matrix up. No row is an opinion: each one shows something already in the operation —the hour the pot goes on, the scale, the trimmed menu, the record of the 9 dishes that left—. All four push dishes with healthy contribution margin and none asks for a discount, because with an average check of 48.000 COP and the goal of lifting operating margin from 8% to 14% in 12 months without touching menu prices, a counter paid for by discounting fixes the video and breaks the month. Row 03 is filmed with the owner talking about his own numbers, never with the platform's name on screen. Illustrative figures; targets get recalculated with the real sales of the following 2 weeks.
Illustrative example generated with a fictional restaurant. Not a promise of results, not a client case.
This is 2 of 3 parts. Behind the wall sits the **defence sheet**: the pre-written answer to each hook's predictable comment, the list of what never gets answered, and the regret zone with the discarded beliefs —health claims, judging the guest, neighbourhood competitors— plus the exact condition that would bring back the only one left waiting.
See the full example — free accountWorks with these AIs
How to use it
Which of your data it uses
This assistant works with 15 of the 23 fields in «My Restaurant», among them Restaurant name, Business type and Cuisine or concept. It does not ask out of curiosity: these are what make the answer speak about YOUR scale instead of an industry average — the same calculation on a business of another size returns a number you cannot decide with. You fill them once and they apply across the library, so the second assistant you open already starts with them in place. Whatever you leave blank, the assistant asks for one thing at a time instead of inventing it, and says so before answering, so you know what it is working from.
Always up to date
Every assistant carries a visible version and date. When an AI changes how it works, the assistant's version goes up and you see what changed.
v1.0.02026-08-20Initial version. Born as the third in the house to work ONE hook type in depth: the curiosity one withholds, the problem-solution one names, and this one argues. The hard part was not writing counters —anyone does that in ten minutes, which is why so many restaurant controversies collapse on their own— but the house-proof filter: a claim that does not stand on a figure from the business is dropped, and dropping is not softening. The regret zone was written after seeing the pattern: the owner does not regret the video, he regrets the answer he typed at midnight.Same method, another variant
Inside the app this is a single control: you press it and the assistant reframes itself, including the variants that don't appear here because they don't have their own page yet.
More assistants in this category
Nothing is invented here: it is translated. An existing hook comes in —yours, another restaurant's, or one fro…
N5 · Complete systemChoose at the top how far to go with what already worked: take the structure apart, adapt its hook to your nic…
N5 · Complete systemDelivers the whole carousel, not just the script: the slide-1 hook with the two discarded ones, the slide coun…
Included in the always-growing library
Access to every published assistant, adapted to your AI and personalised with your restaurant's data, updates included.