Expect a readable pattern in three to six months, with the first genuine sales signal usually appearing between weeks six and twelve, provided the account posts at least three times a week from day one. The bakery account we run went from zero to more than one million organic views in five and a half months, and the industry data explains why the first month of any new account looks quieter than that finish suggests.
This is written for owners, not marketers. Colin ran food and beverage at Generator across 11 properties in 9 countries before Wonder Works, so the yardstick here is covers and average spend, not likes.
Contents
- What should happen in the first 30 days?
- When do the first customers actually arrive?
- Why followers are the wrong number to judge by
- What should you measure at month three?
- What if nothing has moved by month three?
- How an operator should judge it
- When is it fair to pull the plug?
- FAQ
What should happen in the first 30 days?
In the first month a new restaurant account should be posting consistently and watching views from non-followers, not followers or likes. Instagram tests every public post on a small group of people who do not follow you, and only widens that group if the post holds attention. In January 2025 Adam Mosseri, the head of Instagram, named the three signals that decide this: watch time, likes per reach, and sends per reach, with sends (people forwarding your post to a friend) weighted more heavily for reaching strangers. A brand new account has no history for the system to compare a post against, so every post starts from the same small test pool. That is why the first weeks feel like shouting into a cupboard. It is not a verdict on your food. It is the system gathering data.
Two things move the needle in this phase and neither is clever. The first is frequency. Buffer's analysis of 2.1 million posts from 102,000 Instagram accounts, published 13 August 2025, found that posting three to five times a week more than doubled follower growth compared with once or twice, lifted reach per post by around 12 percent, and that weeks with no posts at all pulled accounts below their own baseline. The second is that the posts must be original. Since 30 April 2026 accounts that repost too many of other people's photos and carousels become ineligible for Instagram's recommendations, as reported by Tubefilter that day. Existing followers still see those posts; nobody new is shown them. A restaurant page that runs on reshared food trend content is now invisible to strangers by design.
When do the first customers actually arrive?
The first customers who say "we saw you on Instagram" usually walk in between weeks six and twelve, and they arrive before the numbers look impressive. That timing is not magic. It is the point at which the account has enough posts for the system to have found the ones that travel, and enough of those have been sent between friends for a few strangers to act on them. On the bakery account we manage, 90 percent of 1,133,293 total views came from organic distribution rather than paid promotion, which tells you where the customers came from too: from people who had never followed the page and saw a post a friend sent them.
What the first signal looks like in practice: direct messages asking whether you have a table on Saturday, questions about opening hours in the comments, people tagging a friend under a dish, and, on Google, a rise in direction requests from your Business Profile in the same weeks. If you want the second of those to work, your Google Business Profile has to be complete first, because that is where the social viewer goes to check you are real.
Why followers are the wrong number to judge by
Followers lag views by months and tell you almost nothing about how many people actually saw the food. On the bakery account, total views divided by followers works out at 275 views for every single follower, because the overwhelming majority of people who watched a post never followed the page. They did not need to. They needed to know the bakery existed and where it was.
The industry numbers say the same from the other side. Socialinsider's study of 872,075 brand posts, published 3 September 2026, put Instagram's reach rate at 3.20 percent in August 2026, with the 2026 average at 4.90 percent, down 14 percent year on year. Reach rate is the share of followers a typical post reaches. Pages with 1,000 to 5,000 followers reached 6.65 percent of them, the highest of any size bracket. So a small restaurant page reaches a bigger slice of its own followers than a big brand does, and the real upside sits outside the follower count altogether, in the non-follower views the test-and-widen system hands out to posts that earn it. Judge the account on views and on who those viewers are, not on the follower line.
What should you measure at month three?
At month three you should be able to see four things moving in the same direction: views from non-followers rising month on month, sends per reach above one percent on your better posts, profile visits and direction requests climbing on Google, and a real change in the room on the days you pushed. The first three are in Instagram Insights and your Business Profile, free, and take ten minutes a month to note down. The fourth is your till.
The one figure most owners never open is sends per reach, which is the number of times a post was forwarded divided by the people who saw it. Mosseri named it in January 2025 as one of the three signals that matter most. A post that gets sent is a post doing your word of mouth for you. If your best posts are getting sends and your worst are not, you already know what to make more of.
What if nothing has moved by month three?
If views from non-followers are flat after twelve weeks of consistent posting, the problem is nearly always the content, and it falls into one of four buckets. The first is frequency: fewer than three posts a week and the Buffer data says you are barely above the no-post penalty. The second is reposting: if the page has been running on reshared content since 30 April 2026, it can be excluded from recommendations and no amount of hashtags will fix that. The third is that the posts are about the venue rather than the customer: the new menu, the new terrace, the anniversary, all of which matter to you and none of which a stranger would send to a friend. The fourth is polish: overproduced content is losing to phone footage of real people, and we wrote up the data on that earlier this year.
What paid promotion does and does not do here matters. Boosting a post buys reach for the days you pay for it and then stops. It is useful for putting a specific offer in front of a specific radius on a specific week. It does not fix a content problem, it just shows the content problem to more people faster. We use Meta ads for clients, but only after the organic account is producing posts people send, because that is the cheapest possible test of whether the creative works.
How an operator should judge it
Judge social media the way you judge any other spend in the building: by covers, average spend and which dayparts filled. Running 11 properties across 9 countries taught one thing above all: a marketing number that cannot be tied to a shift on the rota is a number you can safely ignore. So pick your weakest daypart, a Tuesday lunch or a Sunday evening, and give the account a single job for six weeks: fill that. Post the dish, the offer and the room for that slot, and then compare covers for that daypart against the same six weeks the year before. That is a test an owner can read without a marketer in the room.
The reason our F and B partnership starts with dish costing and daypart data rather than a content calendar is exactly this. Social media that fills a room you are already losing money on is not a result. Social media that fills the quiet Tuesday with a dish you make 70 percent margin on is. The content is the easy part once you know which room you are trying to fill and with what.
When is it fair to pull the plug?
It is fair to stop, or to change who is doing it, after ninety days of genuinely consistent, original posting with no rise in non-follower views and no change in messages, comments or direction requests. It is not fair at thirty days, because the test-and-widen system has barely started, and it is not fair if the posting was two a week with a fortnight gap in the middle, because the no-post penalty explains the result on its own. Write down the views from non-followers and the direction requests on day one, look at them again at day ninety, and let those two lines make the decision for you.
If you have been posting for months and the account has never found a post that travels, it can be worth a second opinion before you write off the channel. We are happy to look at an account like this, and the fix is usually one of the four buckets above, not a mystery. And if you are a seasonal venue wondering whether any of this applies in the quiet months, the off season is when the account is built, not when it is paused.
FAQ
How long does it take for a restaurant to get customers from Instagram? The first customers who mention a post usually arrive between weeks six and twelve, and a readable pattern of views and bookings takes three to six months, assuming three or more original posts a week throughout.
How many times a week should a restaurant post on Instagram? At least three and ideally three to five, which Buffer's 2.1 million post study found more than doubles growth compared with one or two, with reach per post rising around 12 percent. Weeks with no posts push an account below its own baseline.
Do followers matter for a restaurant on social media? Much less than views from non-followers. On the bakery account we manage, views work out at 275 for every follower, and 90 percent of 1,133,293 views were organic. The average Instagram post reaches only 3.20 percent of followers anyway, according to Socialinsider's September 2026 data.
What is a good reach rate for a small restaurant Instagram page? Socialinsider's September 2026 benchmark puts pages with 1,000 to 5,000 followers at 6.65 percent reach per post, the highest of any size. Anything above that is doing well. The bigger measure is how many non-followers the post reached beyond that.
Does boosting posts speed up results? It buys reach for the days you pay and then stops. It is useful for a specific offer in a specific area on a specific week. It does not fix content that nobody sends to a friend, so get the organic posts working first and boost the ones that already travel.
When should I give up on a social media manager? After ninety days of consistent, original posting with flat non-follower views and no change in messages or direction requests. Not at thirty days, and not if the posting was inconsistent, because the gaps explain the result on their own.
Sources
- Socialinsider, Social Media Reach Statistics 2026, 3 September 2026: 872,075 posts analysed; Instagram reach rate 3.20 percent in August 2026; 2026 average 4.90 percent, down 14 percent year on year; 6.65 percent for pages with 1,000 to 5,000 followers.
- Buffer, How Often Should You Post on Instagram, 13 August 2025: 2.1 million posts from 102,000 accounts; three to five posts a week more than doubles follower growth and lifts reach per post by around 12 percent; the no-post penalty.
- Dataslayer, Instagram Algorithm: ranking signals Mosseri confirmed, reporting Adam Mosseri's January 2025 statements on watch time, likes per reach and sends per reach.
- Tubefilter, 30 April 2026: accounts that repost too many photos and carousels become ineligible for algorithmic recommendations.
- Wonder Works client data, Kato Edno Vreme bakery: 1,133,293 views, 90 percent organic, 275 views per follower, zero to more than one million organic views in five and a half months.