Impressions, Reach and Views Are Three Different Things, and Only One of Them Is People

A client asked me last spring how many people had seen her pins the month before. I said ninety-two thousand, because that was the biggest number on the screen and I had been reading it that way for years.

The honest answer was about thirty-eight thousand accounts, and fewer actual humans than that. I had been quoting impressions at her as though they were an audience — in a written monthly report, with a chart, for seven months.

She never queried it, which is the part that still nags at me. Nothing I sent her was false. I had simply put the wrong word in front of a true number, and a wrong word in front of a true number is far harder to catch than a lie.

Impressions, reach and monthly views sit within a few centimetres of each other in the dashboard, they rise and fall roughly together, and exactly one of them counts people. Here is what each one is, precisely — and then the part that actually changed how I work.

One sentence each, and then the arguing

The definitions are short enough to be embarrassing, given how much trouble they cause.

• Impressions — the number of times a pin of yours was put on a screen. Not looked at, not read, not considered. Rendered.
• Reach — how many separate accounts that happened to. One person, however many times they scrolled past, counts once.
• Monthly views — a rolling thirty-day impression total across your own pins and pins from any account you have claimed. It is the number people put in their bios.

That is genuinely the whole distinction and it takes about eleven seconds to read. Knowing it and reading the dashboard as though you know it turn out to be two different skills, and I had the first one for years before I had the second.

If you want the rest of the family — saves, pin clicks, engagement rate, the lot — the full glossary covers all of them. This post stays on the top three, because the top three are where nearly all of the confusion lives and where all of the bragging happens.

Impressions measure Pinterest’s behaviour, not your audience

The reframe that fixed this for me is one sentence, and I have said it to enough people now that I can watch the moment it lands.

An impression count tells you how hard Pinterest is currently pushing your pin. It is a statement about the algorithm, not about a crowd.

Read that way it becomes properly useful rather than merely flattering. A pin whose impressions triple in a fortnight has been picked up by distribution — the feed has decided to try it on more screens. That is real news and it is worth acting on. It is simply not the news “three times as many people like my work”.

It also explains the swings that make everybody panic. Impressions can double because one pin landed in a busy feed for four days, and halve again when that stops, without one thing about your account having changed. The mechanics are in how the feed decides what to show, and reading that once takes most of the emotion out of the graph.

There is a longer breakdown of what a single impression does and does not represent in what impressions actually measure, including the awkward business of your own scrolling turning up in there.

Reach is the closest thing to a headcount, and it is still not one

Reach counts accounts. Accounts are not people, and the gap runs in both directions at once.

One person with a phone and a laptop can be two accounts. A shared household account is several people counted once. And an account that was technically shown your pin while its owner scrolled past at speed with their thumb still counts, whether or not a single photon of your work registered in anybody’s brain.

So reach is an upper bound on the number of humans, not a census. It is still comfortably the best of the three, because the one thing it cannot do is inflate itself through repetition — and repetition is precisely what makes impressions so misleading.

One practical wrinkle worth knowing: reach is not offered everywhere impressions are. Depending on which view you are in and which date range you have picked, you may get one number and not the other. That is part of why impressions became the figure everybody quotes — it is the one that is always there.

One person, six impressions, one reach

Here is a week I traced on one of my own pins, because I wanted to see the shape of it rather than argue about it in the abstract.

Monday: it appeared in her home feed while she was scrolling before bed. Wednesday: it came up in a search she ran twice, so twice more. Thursday: it showed underneath a different pin as a related suggestion. Saturday: two more passes in the feed. The following Tuesday she saved it.

Impressions: six. Reach: one. Saves: one. Every one of those numbers is correct, and if you read the first without the second you have invented five people who do not exist.

Six is not even a dramatic example. On a pin the feed likes, in a niche somebody browses often, ten or twelve sightings from one account over a fortnight is entirely ordinary.

And Pinterest re-shows pins on purpose. It behaves much more like a search engine with a long memory than like a timeline — a pin that is still relevant nine months later gets put in front of the same people again, and every one of those is a fresh impression against an account that already knows the pin perfectly well.

That mechanism is behind every “my impressions are enormous and my traffic is nothing” message I get, and it is not a fault in either number. It is one number counting showings and the other counting arrivals.

Monthly views is the bio number, and the one I trust least

Monthly views is impressions with a rolling window and a wider net. It includes pins from accounts you have claimed, which means other people pinning your images can move it, and it rewrites itself daily as the oldest twenty-four hours fall out of the back of the window.

The practical consequence is that it swings hard for reasons that have nothing to do with you. A single pin catching a seasonal wave carries the whole figure up, then drops out of the window a month later and takes it back down — and you spend that second month wondering what you broke. Why monthly views swing so violently walks through the specific causes.

It happened to me with a storage-basket pin one January. The figure roughly doubled, I felt clever for three weeks, and by mid-February it was lower than it had been before the pin existed. I had done nothing in either direction.

What monthly views does not tell you is more interesting than what it does:

• How many people. It is an impression total, so one committed scroller is counted over and over.
• Whether the pins were even yours — pins from a claimed account feed the same number.
• Whether anybody saved anything, clicked anything, or arrived anywhere at all.
• Whether it will still be there in five weeks, which for a seasonal spike it almost certainly will not.

The ratio that finally made these numbers useful to me

Divide impressions by reach. That gives you the average number of times each account saw your pin, and it is the only figure in this whole area I check deliberately rather than glance at.

1. Take both numbers from exactly the same window — the same thirty days, or the same seven. Mixing windows produces a ratio that means nothing, and that was the first mistake I made with it.
2. Divide impressions by reach. Two and a bit means the average account saw the pin roughly twice.
3. Write it down, and do it again next month. The direction is worth far more than the value.

On my accounts it usually sits somewhere between about 1.4 and 2.5. I want to be honest that Pinterest publishes no benchmark for this — it is a number that is useful against your own history and close to meaningless against a stranger’s.

The direction is the good part. A ratio climbing while reach stays flat means Pinterest is showing your work repeatedly to a pool it has already found: you are saturating, not growing. Reach climbing while the ratio holds steady means new people are arriving, which is the only version of growth that eventually produces clicks.

That distinction is completely invisible if you only ever watch impressions, because both of those situations draw the same rising line. It took me an embarrassingly long time to notice that the line I was proudest of was mostly the first one.

The ladder, and why every bio quotes the top rung

Line the numbers up by size and the whole thing becomes obvious. Each rung down is smaller, harder to inflate, and more closely related to somebody actually doing something.

The drop from reach to engagements is where most accounts learn something uncomfortable. The drop from engagements to outbound clicks is where the money either is or is not — and those two are separate events with separate causes. The difference between a pin click and an outbound click is the most diagnostic gap in the entire dashboard.

The ladder is also why a report built on the top rung can look magnificent while the bottom rung is a rounding error. Mine did, for seven months, in a document with my name at the bottom of it.

What I report now, and what I ignore

Reach and saves, every time, in that order. Reach because it is the nearest thing to a headcount I am ever going to be given, and saves because a save is the only event here that costs the person anything.

Impressions I still keep, but only as a trend line and only ever labelled “distribution”. Never as an audience, and never in a sentence containing the word “people”. That is the specific habit I had to break, and writing the label differently was what broke it.

Monthly views I do not report at all. If somebody asks for it I hand it over and say plainly that it is a rolling total which will move for reasons neither of us controls.

Once a quarter I look at who those accounts actually are instead, which is considerably more useful than how many of them there were. And if the reason behind any of this feels arbitrary, what the algorithm is optimising for explains most of the arbitrariness away.

None of these three numbers tells you what to make next, though, which is the only question they exist to serve. Reach tells you how many accounts were reached; it cannot tell you what any of them were looking for. That comes from demand — which phrases people in your niche are already searching, and which pins they are already saving — and it is the whole job PinPinterest does. If you are not even sure your corner of Pinterest is one of the busy ones, the niches we cover is a sensible place to start.

The client took the correction well, for the record. Her reach was a third of the number I had been sending her, and her traffic made a great deal more sense once we stopped measuring it against a figure that had never been a crowd.