A café posts a video and it does well. Forty thousand views, comments from four countries, a nice week. The room is just as quiet on Tuesday.

Nothing went wrong with the video. It reached people who cannot come. In this business the only viewer who matters is one who is nearby, or about to be, and everything worth doing follows from that sentence.

Reach is the wrong number

A restaurant does not have a distribution problem, it has a door. Someone has to physically arrive, usually within a short walk, usually within a few days of seeing the video.

That makes a small local audience more valuable than a large distant one, and it makes the usual measures actively misleading. A piece with three hundred views from the surrounding streets can fill a service. A piece with fifty thousand from elsewhere fills nothing and feels like success, which is the expensive part.

Two customers who never overlap

Almost every place in this country serves both, in different proportions, and one content line cannot speak to both.

The visitor decides in ninety seconds

They are standing outside, or they are on a phone two streets away, and they are choosing between you and four others. They cannot read the menu culture, they do not know what a normal price is here, and they are afraid of the tourist trap.

What settles it for them is legible information: the actual prices, the actual dishes, the room with real people in it. A moody close up of a plate tells them nothing they need.

The neighbourhood decides by habit

Someone who lives nearby is not choosing a restaurant, they are deciding whether to cook. They already know you exist. What moves them is a reason today: a dish that is only on this week, the fact that you are open when they assumed you were not, a table free at eight.

January is the test

The two audiences have opposite seasons, and that is the part that decides whether a place survives. Somewhere living on visitors is empty in the quiet months, and somewhere living on locals is squeezed out in the busy ones. Content made in August for visitors does nothing in January, so the neighbourhood work has to be done before it is needed.

Who is decidingWhere they are watchingWhat gets them through the door
A visitor outsidePhone, on the street, right nowVisible prices and a full room
A visitor planningSaved video, two days aheadThe walk from a landmark, and the opening hours
A neighbourLate afternoon, at homeA reason that only applies today
A regularAnywhereSomeone they recognise behind the counter
A group organiserMessages, in advanceWhether ten people actually fit

The last row is the one restaurants never film and get asked about constantly. Thirty seconds showing the back room and saying how many people fit answers a question that otherwise costs three messages and often ends in a booking somewhere else.

What to film, and what is pointless

The single dish, beautifully lit, is the most commissioned and least useful piece in this vertical. Every competitor has one, they all look the same, and none of them tells anyone whether to come.

What works is unglamorous and specific. The room at the hour it is actually busy, with the noise left in. The person who cooks, saying one true sentence about one dish. The door filmed from the street, so someone arriving recognises it. A menu board where the prices can be read without pausing.

And the thing nobody does: film the walk. Sixty seconds from the nearest well known corner to your door, in real time. It removes the exact friction that makes a visitor give up, and it costs one afternoon a year.

Who should film it

A creator who lives nearby, and preferably one who will come back. The value here is not production quality, it is repetition and locality: the same person, the same place, at different hours across a season, builds something that reads as a real address rather than a campaign.

It also solves the language question by accident. A local creator posting in Portuguese reaches the neighbourhood, and the same footage with a different voice reaches the visitor, from one shoot.