Hiring mistakes with conferencistas de ingeniería de Menú para el Crecimiento de Restaurantes — and the criteria that work

Step 1: ask for the matrix before the reel

Demand one deliverable before anything else: a profitability-popularity matrix from a real menu the speaker reworked, with contribution margin per dish and twelve months of sales mix. That Bogotá committee dropped its three 2026 finalists precisely there, and all three arrived with video testimonials. Verify the sheet with a finger on it — twenty items minimum, every cell carrying food cost percentage, selling price and units sold. Four quadrants with dish names and no figures mean the candidate read the 1982 Kasavana and Smith paper; working a menu is something else. At RadarSpeakers the spreadsheet gets read before the stage presence. THRESHOLD: under 70% of cells with a number, send the brief back and ask for the full version.

Step 2: write a brief with YOUR operation's numbers

80% of a keynote's quality comes out of the brief you draft. Almost no organizer writes it properly, and for years I gave it less weight than it deserves. Send the speaker three hard data points about your audience before signing: average check for the segment, the weight of off-premise business, around 75% of traffic according to Circana, and the food cost range your attendees report. An expert comes back with uncomfortable questions inside 48 hours; the average one replies that this is plenty. Two pages is the format — market context, three prioritized pain points, the event's definition of success — and the audit is arithmetic: count how many of your figures survive into the outline they send you. Fewer than five in a sixty-minute keynote means you are getting the usual talk with your logo on slide one.

Step 3: separate good from excellent by the second axis

One axis separates good from excellent on this topic. The average speaker stays with stars, plowhorses, puzzles and dogs, content your attendee reads free in any hospitality management textbook. Cross margin and popularity against labor cost per dish inside the kitchen station and you are in another league: AI-driven scheduling cuts labor cost by 8% to 12%, according to TimeForge (2025), so that axis moves real money. Above them sit the few who also weigh table turn speed and channel mix, because when 37% of adults order delivery at least once a week (UpMenu, 2024), the dish worth pushing changes. Ask in writing which variables they use beyond the original two. Fewer than three extra and you are looking at an average profile, whatever the fee says.

Step 4: demand live analysis, not a generic slide

The cheapest filter in the whole process costs nothing extra: ask them to redesign an audience menu on stage. You hand over two anonymized menus from volunteer attendees seventy-two hours ahead, and the speaker leaves the platform with two dishes recosted in front of the room — which price rises, which garnish goes, what margin the house gains. Suppose you skip it and take a recorded case study instead. The room still applauds, the event report still looks good, and ninety days later no manager has touched a menu; you bought morale, not an intervention. Video gets rehearsed, somebody else's menu does not. Minute the exercise with contribution margin before and after each dish touched: under 3 percentage points of improvement on the analyzed mix, you paid for theater with data.

Step 5: fee, rider and honorarium with no ambiguity

90% of logistics conflicts start with a one-line quote, so close the fee itemized before you book the auditorium. Four separate lines, none of them optional: keynote honorarium, hours of content customization, travel with flight class and hotel nights, post-event materials. The technical rider travels apart and in writing — screen, clicker, headset microphone, internet for the live exercise — because on load-in day nobody remembers a phone conversation. Booking direct, with no intermediary commission, is what frees the money to negotiate those preparation hours a bureau usually bundles away; at RadarSpeakers the profiles publish fee ranges and 2026 calendar availability on the same page. If the contract does not say how many preparation hours it includes, it includes none.

The four mistakes that keep repeating

Hiring on fame remains the most expensive mistake in the catalogue, and it is only the first. Next comes accepting the stock keynote without demanding adaptation, when a serious speaker spends between 6 and 10 hours rebuilding content for a hospitality congress. Format blindness does equal damage: a 45-minute plenary for 800 people cannot carry the same matrix exercise as a workshop of 25 managers, and whoever never asks about capacity will not ask about objectives either. The fourth is the quietest, because nobody misses it until the contract is signed: verifiable references from previous organizers. Two phone numbers, not two logos. One piece of context almost nobody uses when building the panel: ACODRES reported a 9.8% rise in dish prices across Colombia in 2025. Your audience walks in with menus already touched and smells an improviser.

Step 7: post-event materials with shelf life

What survives the applause decides whether the investment paid, and here the excellent separates from the good with brutal clarity. Three pieces into the contract: the matrix template as a spreadsheet, ready for each attendee to load their own menu, a two-page document with the decision criteria and a 30-day asynchronous question window. The average speaker sends a PDF of the slides, which nobody opens twice. That template has to calculate contribution margin, popularity index and cutoff point on its own, with no formula knowledge required from the manager. At 45 days, count how many attendees returned a filled matrix: under 15%, either the tool was too complex or the talk never explained how to use it.

Closing checklist: how to know it went well

It went well if you can tick six boxes with evidence in hand rather than hallway impressions. The speaker's matrix, twenty items or more, a figure in every cell. Five data points of yours, minimum, inside the final outline. A record of the live redesign with margin before and after. A contract itemizing honorarium, customization hours, travel expenses and materials. Two phone references from organizers who already booked them. And audience NPS above 60, measured as the room empties, never by email three days later. Fail the first box and the other five will not save you: you are buying reputation. Start today with the simplest move, write the email asking for the matrix of a menu they actually reworked and see who answers with a file attached.

Four differences that separate good from excellent

Kasavana and Smith published the profitability and popularity matrix in 1982, and the AVERAGE speaker stops right there: four labels — stars, plowhorses, puzzles, dogs — one textbook recommendation, applause. Accurate and useless at the same time. That material sits in any hospitality management manual, and your attendee reads it free in twenty minutes. You recognise the GOOD one by the axis almost nobody works: what the labour on that dish costs inside its kitchen station, how fast the table turns and the weight of off-premise business. Using the numbers the organiser handed over during the brief, this speaker dissects a menu from the room live, then recosts two attendee dishes in front of everyone before leaving the stage. The EXCELLENT one commits. They name the dish that has to leave the menu and hold the room's discomfort, they admit to years of recommending price rises before fixing the recipe card —exactly backwards— and they leave a 90-day numeric checkpoint the organiser can audit without a phone call.

Four differences that separate good from excellent — in practice

That is also where the tension nobody resolves on stage shows up: the most profitable dish is rarely the one the kitchen loves, and a new menu without the chef alongside sabotages itself inside three weeks. The way out is not imposition, it is costing the dish with the chef at the table, because a number argued in the station survives a busy service. One more line divides operators from theorists, and it runs through paper: a serious speaker NEVER recommends dropping the printed menu for a QR code alone. Print sets service rhythm and carries suggestive selling, along with the story of the menu; the QR adds delivery, accessibility, price updates and analytics. Selling the fully digital menu from a 2026 stage confuses a channel with the guest experience, and that question fits neatly into the screening interview.