Instagram marketing for restaurants: 3 playbooks

Instagram marketing for a restaurant comes down to three playbooks: the plate feed, the room feed and the offer feed. Each converts a different diner, lands on a different part of Instagram and costs a different number of hours a week. Most restaurants run one by default, nearly always the plate feed. The useful question is not which playbook wins. It is what ratio of the three fits the way people actually book your tables.

What are the three playbooks?

The plate feed. Every post is a dish, shot at the pass or at the table, with a caption that names it and invites you in. It is the default because the material is at hand at every service.

The room feed. The place and the people: the dining room at seven, the regular who has had the same table for eleven years, the fishmonger at dawn. The dish appears as context, not as subject.

The offer feed. A reason to come this week: the Thursday special, the seasonal menu going live, the last tables for New Year's Eve. Each post carries a date and a way to book.

Open ten restaurant profiles in your city: the usual ratio is close to eight plates, one room, one offer. Four criteria show why that is the problem.

Which diner does each playbook convert?

A restaurant has two readers on Instagram, arriving from opposite directions.

The stranger comes from outside the app: a friend's story, the Instagram link on your Google Maps listing, a location tag. They land on your profile and read your last nine posts as a menu and a vibe check.

The regular already follows you and never opens your profile. Your post has to reach them in a crowded feed and give them a reason to come back this week rather than eventually.

The plate feed converts the stranger. Nine plates answer the question they came with: what will I eat, and is it worth the trip. It does almost nothing for the regular, who has eaten the burrata.

The room feed converts the regular, and the friend of a regular. Someone who recognises the dining room, or themselves at the bar, shares the post. The stranger reads a grid of rooms, finds it warm, and leaves to look for the food elsewhere.

The offer feed converts whoever is deciding right now. It is the only playbook where a post can be traced to a booking.

So the grid belongs to the plate, the feed to the room, the booking message to the offer. And since the grid is what a stranger reads first, which posts sit next to each other is a decision of its own.

“

The grid belongs to the plate, the feed to the room, the booking message to the offer. Your booking path sets the ratio.

”

What does Instagram do with each post?

Instagram ranks a post separately for the feed, Explore, search and your profile, and each playbook lives on a different surface.

A plate photo collects saves, and saves travel: to your location page, and to Explore, where a stranger who follows three other restaurants in your city gets shown yours.

A room photo collects shares and tags. Instagram has said publicly, more than once, that a post sent in a private message is one of its strongest signals. The room feed puts you in front of the people who eat with your customers, the only new audience that matters for 30 covers.

An offer post collects profile taps and messages, and not much else. Instagram shows it mostly to people who already interact with you.

One mechanism most guides skip: the Tagged tab, which the stranger opens right after the grid. People tag a restaurant when they post their evening, and what they post is the table, the friends, the room. A grid that looks like their evening gets tagged. A grid that looks like a catalogue does not.

What does each playbook cost a one-owner restaurant?

The plate feed is cheap per post and expensive over a season. The photo takes a minute at the pass. The caption takes ten, because the fifth burrata caption is harder than the first. And the runway is finite. A menu of 30 dishes with three honest angles each, the plate, the ingredient, the moment it gets eaten, is 90 posts: 22 weeks at four a week. The feed reads as repetitive long before post 90, because the sentence shape repeats before the dishes do: adjective, dish, invitation.

The thirty-second test doubles as the mechanism inside any generator you hand the week to. Paste your menu into an AI and ask for ten Instagram posts. Count how many open with an adjective. A generator that only has a menu has one sentence available, and it writes it ten times, then ninety. Give it the sourcing, the team, the hours and the story of the place, and the plate posts start carrying a reason: who caught it, why it is on this week, which table it is best eaten at. A generator that starts from your website reads the about page in the same pass as the menu, and that is the whole difference.

The room feed costs attention, not hours. Every service produces the material; what it costs is someone remembering to capture it when nobody has a free hand. Its runway is unlimited, because it belongs to the family of post ideas whose material the work produces on its own.

The offer feed costs the least to write and the most to plan. The post is a date, a dish and a booking link, but it needs a calendar to exist. The posts that fill a counter answer the question "when can I get it", and that question needs something scheduled behind it.

Running all three at four posts a week, alone, between two services, is where most owners fall back to one playbook. It is also where generating Instagram posts tuned to your restaurant stops being a nice idea and becomes the only way the ratio survives.

Which ratio fits your restaurant?

The ratio follows the booking path, not taste.

Walk-in neighbourhood place, 30 covers, a menu that changes. Grid first: five plates, three rooms, two offers out of ten, with the two most recent posts a plate and a room.

Reservation-led, weekends booked two weeks out. The decision is made days ahead, on a recommendation, so shares matter more than the grid. Four rooms, four plates, two offers about dates. Turn on the reservation button Instagram gives business profiles through its booking partners, and stop writing "message us to book".

Lunch counter or takeaway. The decision is made an hour before lunch. Six offers, three plates, one room, and the offer post goes up before that decision, every day, special or not.

Hotel restaurant. Guests upstairs eat wherever is easiest; locals never think of a hotel as a place to eat. The room feed fixes the second problem, because it shows a dining room with locals in it. A boutique hotel that rebuilt its feed from phone photos taken between two services is the closest published case. Four rooms, four plates, two offers aimed at the neighbourhood.

The ratio that fits you is the one you can hold in the week the walk-in fridge dies on a Friday. A ratio you cannot hold on a bad week is a ratio you do not have.

Open your own grid now and count the last ten posts by playbook. If the count is nine plates and one offer, you already know which diner you have been talking to, and which one you have not.

FAQ

How often should a restaurant post on Instagram? Three to four feed posts a week keeps the grid current. The ratio between plate, room and offer matters more than the number.

Do hashtags still matter for a restaurant? As a search signal, a little. A few tags naming the dish, the neighbourhood and the cuisine help; a block of thirty adds no reach. The location tag does more than any hashtag.

Should I repost photos diners tag me in? Yes, with permission and credit, and prefer the ones that show the room and the evening over the plate. They cost nothing to make.

Do I need professional food photos? For the plate feed, a phone at the pass in daylight beats a quarterly shoot that goes stale by week three.