Restaurant kitchen videos for social media: mistakes vs the right method

Which dish should you film first?

Film first the dish with the best contribution margin that the kitchen can plate fast at peak hour, not the most photogenic item on the menu.

The deliverable for this step is a short list of three dishes chosen with the costing sheet open, each with its food cost below the 32 % ceiling set by the method, plus one line explaining why that dish and not another. If a dish goes viral with a high direct cost, the restaurant ends up selling MORE of what leaves it the least, and the hot line gets swamped with a product that barely pays for itself. For example, if your tenderloin costs 12 and sells for 30, while the house rice costs 4 and sells for 24, the rice leaves more per plate served even though it looks humbler on camera, and it deserves the first video. Checking is easy: anyone on the team can defend the choice with the sheet in hand.

The short script every video needs

Every video needs a four-line written script before filming: the hook in the first seconds, the process the guest never sees, the finished plate and the one action you ask the viewer to take. Sprout Social (2026) found that 52 % of social media users are more likely to engage with short-form video under 60 seconds on Instagram, so the script is written to run under a minute and everything that does not push toward the final action gets cut. The hook is not the pretty plate, it is the question the diner never asked, like why the broth simmers for twelve hours or what the cook does with the fish bones. And the action must be ONE, book for Friday or order by message, because two calls to action cancel each other out. The deliverable is a sheet with the script for each chosen dish, reviewed by whoever signs off on costs.

How to film on a phone without looking like a supermarket ad?

Film on the restaurant's phone, with side light from a window or a lamp, a tight shot on the hands and the real sound of the griddle, and drop any production that erases the kitchen.

There is a tension here that many owners resolve badly: polished aesthetics sell, yet agency-style production kills what the guest actually wants to see, which is the craft. Toast (2024) reports that 84 % prefer to see food and drink photos on a restaurant's social channels, so the plate rules, and the bridge is to care for light and framing while leaving the cook's gesture untouched. I would always add on-screen text, because a large share of the audience watches muted on the bus or at the office. The deliverable for this step is three raw takes per dish, shot in a single forty-minute session before service, with the text and the cut checked by a second person.

When should restaurant cooking videos be posted?

Post them at the moment the guest decides the plan, which usually comes one or two days before the visit, not at the hour the dish leaves the kitchen.

A business lunch gets decided that same morning, an anniversary dinner is planned several days out, and each occasion's video has its own window. The audience is there: DataReportal puts social media user identities in Mexico at 74,9 % of the population (October 2025), and in Colombia at 69,2 % (January 2025). Among younger diners it weighs even more, since 67 % of Gen Z have used social media to decide which restaurant to go to, according to TouchBistro's 2025 report. The deliverable is a four-week calendar that pairs each video with its consumption moment and with the dish the kitchen has ready for that service, never posting what is not on the line that day.

Measure at the register, not in views

A video is measured by bookings, messages and orders of the pushed dish against the baseline recorded before starting, never by views or hearts. Each piece carries a trackable signal: a keyword to order by message, its own booking link or the dish name the server notes when the guest says they saw it on social. What would happen if you measured reach alone? The most-viewed video would look like the winner, the owner would ask for more of that style, the team would film the flashiest dish and, six weeks later, average food cost would have climbed with nobody knowing why. Keep in mind too that 62 % of consumers discover restaurants via Google, more than through Yelp or social (PYMNTS, 2023), so the business's Google profile must show the same dish under the same name. The deliverable is a weekly sheet with those register columns, filled in every Monday.

The mistakes that repeat most when following this guide

The most repeated mistake is filming without having decided which dish to push or what the viewer should do, and nearly every other error grows from it. Next come posting without on-screen text, changing style every week before there is any data, and leaving comments unanswered, which is exactly where trust breaks: 25 % of diners would avoid a restaurant because of criticism on social media, per TouchBistro Diner Trends 2025 as compiled by Tablein. A less visible failure is when all the content comes from the kitchen and none from the dining room, even though in the US 31 % of consumers watch videos of ordinary people on social media when researching local businesses (BrightLocal, 2025), and a guest describing their dish persuades better than a chef presenting it. I got this wrong for years, convinced the chef's video was enough. The fix is cheap: ask a regular for permission and film half a minute at their table.

Closing checklist: how to know everything is done right

Everything is done right when anyone on the team can open one folder and find, without asking, the dish list ranked by margin, the scripts, the published videos with their dates and the weekly results sheet. The review we use at Masterestaurant has five questions answered yes or no: each video's dish sits below the method's cost ceiling, the video runs under a minute with on-screen text, it asks for a single action, it went out in the decision window of its consumption moment, and it carries a trackable signal at the register. If any answer is no, that video does not count when judging the strategy. Diego F. Parra adds a tougher criterion: after six weeks, the pushed dish must weigh more in the sales mix than it did at the start, and if it does not, you change the dish before you change the format. That sales-mix comparison is what you put on the calendar today.

What changes between filming for the sake of it and the right method?

The first thing I correct in any account is a widespread assumption: that the dish you show most should be the most photogenic. It's the other way around.

A kitchen video is a menu engineering lever dressed up as content, and if the plate that goes viral runs above 32% food cost, the restaurant ends up selling MORE of what earns it the least. Choose by contribution margin first, then work out how to make it shine, which side light and a tight frame almost always solve. There is a real tension between polished aesthetics and authentic cooking, and many owners resolve it badly in one direction: ad-agency footage nobody believes, or shaky clips nobody understands. The way out is to separate what you film from what you edit. The process stays raw, with the sizzle of the flat-top and the team's real hands, while the care goes into what viewers won't forgive, which is the opening, the sound and the on-screen text. Here is a nuance most social media for restaurant marketing plans miss.

What changes between filming for the sake of it and the right method — in practice?

PYMNTS (2023) reports that 62% of consumers discover restaurants through Google, more than through Yelp or social, so a video that never reaches the restaurant's Google Business profile is working at half speed.

What happens if a restaurant posts daily with no script for a whole quarter? The kitchen tires by week three because filming eats into prep, shot quality drops, retention falls and the platform shows each new clip to fewer people, so by quarter's end the owner concludes social media doesn't work, when the real failure was cadence. Three scripted videos a week outperform seven improvised ones, and they last. In the Masterestaurant method, Diego F. Parra treats restaurant branding as a consequence of operations rather than makeup on top of them. The brand is what guests confirm on the plate, and the kitchen video is advance proof of that promise, so the question we ask before filming is what we want guests to order next time, and how much it leaves us when they do.