Reach counts distinct people rather than times shown, and it answers mostly one question: how many new people a piece touched. That is why it can climb on a weak day, when something travelled far and landed badly, and fall on a good one, when your regulars engaged heavily. It is read next to impressions and engagement, never alone.
People and times are different units
A LinkedIn post can close the week with 48 impressions and 21 people reached. Both figures describe the same post and they look nothing alike: it was shown 48 times to 21 distinct people, a bit more than twice each. Looking at only one of those columns leaves out half of what happened, and which half you lose depends on which column you kept. Impressions alone cannot tell a piece that convinced two hundred people once from one that chased twenty people ten times.
Impressions divided by reach gives frequency, which is how many times the average person who saw the piece saw it. There is no correct number for that division, and be suspicious of anyone who offers you one: on a small, loyal account it rises by itself, and on a piece that travelled to strangers it falls by itself. What does mean something is its trend. If your frequency has been climbing for three months while reach stays flat, you are talking to the same people more often.
That is the reading behind this guide's headline. A day when a piece gets distributed outside your followers pushes reach up and usually sinks the engagement rate, because it lands in front of people who do not know you and who scroll past. A day when you post something your community loves does the opposite. Neither day is good or bad in itself, and that is the whole lesson.
Which figure each network gives you
Not every network speaks this language. Three of the six the dashboard can measure report reach per post, and the other three do not have it at all, so there you work with what does exist and you say which one it is.
| Network | Reach (people) | Impressions (times) | Plays |
|---|---|---|---|
| Yes | Yes | Yes, on reels and video | |
| Yes | Yes | Yes | |
| Yes | Yes | No | |
| TikTok | No | No | Yes |
| YouTube | No | No | Yes |
| Google Business | No | Listing appearances | No |
The empty cells in that table are not an integration gap: they are metrics each platform's API does not return to anybody. TikTok exposes no impressions and no de-duplicated reach, and its audience figure is plays. YouTube is the same, and what it has is plays and watch time. Google Business measures something else entirely, how many people found the listing on Maps and in Search, which is why it is read in its own section rather than in the same column as a reel.
From there comes the first practical rule of a report: when you compare networks, say what you are comparing. A chart that puts Instagram reach next to TikTok plays is not miscalculated, it is mislabelled, and the client reading it will add two different things together in their head.
Three things you cannot do with reach
Three operations look innocent and break the figure. The first and by far the most common is adding days up. Reach counts unique people, and somebody who saw you on Monday and on Tuesday is one person in the real total and two in the sum. A week of seven daily values of 800 does not reach 5,600 people; it reaches considerably fewer, and the more loyal your audience the fewer it is. Only the platform knows a period's total, because only it knows who repeats, so you either ask it or you leave the cell blank.
The second is averaging it. A mean of daily reach figures is a number with no referent: it is neither the average reach of a person nor that of a post, it is the average of thirty counts that overlap each other.
The third is adding networks together. Instagram reach plus Facebook reach is not people reached across Meta, because the same person can be in both and nobody will tell you how often that happens. If you need one global figure, the honest one is total impressions, which does add up, stated under its own name.
The same rule applies to accounts engaged, which Instagram serves at account level and which also counts unique people. If a metric hands you people rather than actions, do not add it across days even though the spreadsheet will let you.
Why yesterday's number changes tomorrow
Reach is not finished the next morning. Instagram and Facebook can take up to 48 hours to close a day and they revise upwards what they already gave you. YouTube finalises its data two or three days late, to the point that asking it about yesterday returns zeros rather than an error. Google Business also arrives two or three days behind.
The consequence is operational and worth deciding once: a monthly report is not closed on the 1st, it is closed a few days later. Publish it the same day and you will be comparing a finished month against one that is missing its tail, and the new month will always look worse. How that close gets built so that it does not depend on somebody remembering is on the reporting page.
There is a second reason your report and the network's own app never quite agree, and it is not a mistake in either: the windows are not the same. The app shows you the last rolling 7 or 30 days and your report runs from the 1st to the 31st, and reach, not being summable, cannot be trimmed from one window to the other.
Everything in this article happens in one place in GoFeed.
Try it freeWhat to compare against, given there is no benchmark
The next question is always the same: how much reach is good. There is no publishable answer. It depends on the network, the size of the account, the sector and the format, and any round percentage doing the rounds was invented somewhere without a source. An account with 2,000 followers in a narrow niche and one with 2,000 followers for a local shop have nothing in common.
What is comparable is your own account against itself, and there are three ways to do it properly. The first is the same period of the previous month with the same number of posts, because if you published twice as much, reach rose on volume rather than on quality. The second is the same network and the same format: a reel and a carousel are not in the same league and are not distributed alike. The third is the median of your last ten pieces rather than the mean, since a single runaway post drags the average with it.
And when somebody outside asks for a reference number, the one you can give without lying is your own: average reach per post over your previous quarter, quoted together with how many pieces sat behind it.
What to look at when reach falls
Falling reach is not a diagnosis, it is a symptom, and there is a sensible order for finding the cause. First, volume: fewer pieces almost always means fewer people, and it takes a minute to check. Second, the format mix, because a month with more carousels and less video moves reach without anything being wrong. Third, engagement, which tells you whether the problem is distribution or content: reach down with engagement rate up is a lower volume month, and reach down with the rate down too is something else.
That rate is calculated from the period's totals and not as an average of the daily rates, which gives a different and worse number. The formula, its traps and what belongs in the divisor on each network are in the engagement rate guide, and if you only want the number, the engagement rate calculator does it with the data already in front of you.
If after those three checks reach keeps sliding with volume and formats steady, the problem is not the month, it is the account, and that is what the stalled account guide is for. The daily series, the per period de-duplication and the per network breakdown are kept by the analytics module, which is where you see whether what fell was the people or the times.