Facebook Reach vs Impressions: One Counts People, One Counts Screens

Open Ads Manager and hover the column headers. Meta's own definitions are short and they are the whole story: reach is the number of people who saw your ads at least once, impressions are the number of times your ads were on screen. Same content, two different units. One is a headcount, the other is a delivery count.

The consequence shows up the moment you try to do arithmetic with them. Impressions add up. Run a post for seven days and the week's impressions are the sum of the daily ones. Reach does not work that way. A person who saw your ad on Monday and again on Thursday is two impressions and one person, so the week's reach is smaller than Monday plus Tuesday plus the rest, and the longer the window, the wider that gap grows.

The same trap sits across campaigns. You cannot add the reach of two ad sets together and call it your audience. Meta deduplicates inside a reporting scope; two scopes stapled together in a spreadsheet get no such treatment. Their overlap is invisible to you, so the combined figure sits somewhere between the larger of the two and their sum, and the report will not tell you where. Pull it at the level that contains both ad sets, or you are guessing.

Facebook counts an impression when your content is on screen. That is the entire bar: no reading required, no dwell time. Someone thumbing past at speed adds an impression. So does that same person scrolling back up to the ad an hour later. None of it adds anybody to reach.

That is also why the gap between the two numbers widens when your audience is small. Narrow the targeting hard enough and delivery keeps returning to the same people, because there is nobody else to show it to. The impression count climbs, the reach line flattens, and the campaign looks busy while the room stays the same size.

Placement plays into it too. Full-screen vertical surfaces are built for continuous scrolling and replay, so the same account passes the same content more often there than it would a square post sitting once in a crowded feed. That is a reach-versus-impressions difference before it is a creative one, and part of why vertical format changes how far a Facebook post travels rather than just how it looks.

Impressions divided by reach is frequency, and Ads Manager carries it as its own column. Some column presets list it already; if yours does not, add it through Columns, Customize Columns, where it sits with the performance metrics. It reads as an average: how many times the typical person in your reach saw the thing.

A frequency of 1.2 means most people saw it once. A frequency of 6 over a two-week flight means you are talking to a small group repeatedly, and whether that is good or catastrophic depends entirely on the offer. A brand campaign can absorb repetition. A single hard-sell creative usually cannot, and the tell is a cost line that drifts upward while reach stops moving, since you are paying for impressions on people who already decided about the ad.

Meta publishes no target frequency, and no single figure in the interface is marked as too high. The useful practice is to watch your own frequency against your own cost per result over the same window and find the point where one starts rising as the other does.

Reach is deduplicated by account, and an account is not a person. One human with a personal profile and a second login is two. A tablet shared by a household is one. Meta has long labelled reach an estimated metric rather than an exact tally, and says metrics that count people are calculated from sampled data. That is worth knowing before anyone builds a slide on the last two decimals.

Organic reach carries a heavier problem. A high reach number on a Page post says the post was distributed. It says nothing about whether the people distributed to were the ones you wanted, which is the same gap that makes follower counts such weak evidence of an audience. Reach borrowed from a comment thread that went sideways is still reach. Those people never went looking for the Page; the thread put you in front of them, and the column records no difference between them and a long-time follower.

There is one more edge. Chasing reach through engagement prompts has a documented downside on Facebook, since the platform has publicly demoted content that asks for interactions as a shortcut, and the fallout from like bait landing on reach and followers is slower to undo than it was to cause.

Impressions are the denominator in CPM, the cost per thousand impressions, which means every CPM comparison is a comparison of delivery cost, not of people cost. Two campaigns with identical CPMs can reach very different numbers of humans if one runs at frequency 1.4 and the other at 5. If you care how many people heard you, CPM answers the wrong question, and cost per reached person is a figure you have to work out yourself.

The other soft spot is the word itself. Meta has been consolidating public-facing language around "views" across Facebook and Instagram, so depending on the surface and the date, the number in front of you may be labelled views, impressions, or plays, with different counting rules behind each. That relabelling landed on organic content first; Ads Manager still reports impressions for ads. Video views have their own definitions tied to playback and are not interchangeable with impressions on a photo post. Read the tooltip before you compare two numbers from two different screens.

Alone, neither column is worth much. Together they describe a shape.

Organic posts get the same treatment with fewer levers. Page reach that slides while your posting rate holds steady usually means distribution is being allocated elsewhere, and the first thing to check is whether engagement per reached person moved, since the connection between likes and organic reach is where that pressure shows up first. If the deeper problem is that too few people ever land on the post to begin with, the distribution side is a separate job from the creative side, and our Facebook growth services page sets out what that covers.

Practical step for this week: add Frequency and Reach as columns next to Impressions in Ads Manager, set the date preset to Maximum rather than Last 7 Days (Meta retired the older Lifetime preset, and reporting reaches back 37 months), and write down the frequency for every active ad set. Then do it again next Monday. A single frequency figure is trivia. Two of them, a week apart, tell you whether you are still reaching anyone new or just re-showing the same ad to the same room.

Neither, because they answer different questions. Reach tells you how many people you touched, impressions tell you how many times content was delivered. For awareness goals where breadth matters, reach is the number to optimize. Under the Awareness objective you can set the performance goal to maximise reach, and reach-goal ad sets carry frequency controls for how often one person is served. For direct response, repeated exposure is often the point, so impressions and frequency matter more. Track both and read them as a ratio.

That is CPM, cost per thousand impressions, and Facebook does not publish a fixed rate because there is not one. CPM is set at auction and moves with your country, audience, placement, objective, competition and season, so the same ad can cost materially more in December than in February. The only CPM that means anything to you is the one in your own Ads Manager for your own targeting.

There is no published benchmark, and any number quoted as a standard was invented by whoever quoted it. Impressions scale with budget and audience size, so a large number proves spend rather than performance. Judge impressions against your own history and against frequency instead. If impressions doubled while reach held still, you bought repetition, not audience, whatever the headline figure says.