Pinterest Analytics for a Small Shop: Five Numbers and What to Do With Each
About Alex: a pen name. There is no shop behind it, and no results are claimed. What follows is quotes from Pinterest's own help pages, arithmetic on invented numbers, and my opinion, and I'll mark which is which.
In short
Pinterest analytics lists many metrics. For a small shop I would read five: impressions, saves, Pin clicks, outbound clicks and the outbound click rate. Each one answers one question, and none of them is a sale. Read them for the same date range as last period, name the pattern, and write one decision. (Picking these five is my judgement; the definitions are Pinterest's.)
What you get from reading it:
- A one-page reading sheet for the five numbers
- A worked example for an imaginary account, every rate recomputed
- One decision for each common pattern, plus a list of misreadings
You post a few Pins, wait, and open analytics. There is a graph, a row of tiles, a list of top Pins, and a filter icon that opens a dozen more choices. Some numbers are going up and some are going down, and you cannot tell whether any of it matters. So you close the tab, or you decide that the one number that moved is the one that counts.
I would slow down at that point. A small shop needs a short list of numbers, a way to compare this period with the last, and a rule for what to do next.
Who this is for. A seller with a Pinterest business account and at least a few Pins that link to a page of their own. Pinterest's help page says you need a business account to view Pinterest Analytics [A1]. If your Pins do not link anywhere yet, the page they point at comes first: How to Write Your First Product Page: Seven Blocks, in Order.
What it does not do. It gives no benchmark. I have no source for a good save rate, a good click rate or a good impression count, so none appears here. Your own earlier numbers are the only comparison I can defend.
One imaginary seller, Rin, carries the example. Rin, the account and every number attached to it are invented to show the arithmetic. They are not results and not typical of anything.
What the analytics page actually counts
You'll end up knowing which words on the screen mean what, in Pinterest's own wording.
On 26 October 2026 I fetched Pinterest Business help, "Review Pinterest Analytics" [A1]. The page says the Overview shows your overall performance, performance over time, top boards and top Pins from the last 30 days, and that "these real-time metrics are estimates and subject to change. As a result, they might decrease."
Its metrics table defines the terms this guide uses. Quoted, with the page's own wording:
- Impressions: "The number of times your Pins or ads were on screen."
- Saves: "The number of times people saved your Pin to a board."
- Pin clicks: "The total number of clicks on your Pin or ad so it opens in closeup."
- Outbound clicks: "The number of times people perform actions that lead them to a destination off Pinterest." A note adds: "Outbound clicks are measured starting from the date the link is added."
- Outbound click rate: "The total number of clicks to the destination URL associated with your Pin divided by the total number of times your Pins were on screen."
The page lists impressions, saves, Pin clicks, follows and outbound clicks as the core metrics for every Pin format [A1]. It lists engagements as saves, Pin clicks, outbound clicks, carousel card swipes and secondary creative clicks.
One wording problem. Pinterest's separate "Review your Pin stats" page, which I fetched the same day [A2], defines Pin clicks differently: "The total number of times people tap on a Pin or ad that leads them to a destination on or off of Pinterest." It defines outbound clicks as "The total number of clicks to the destination URL associated with your Pin or ad."
Neither page explains how the two wordings fit together, and I did not test it. Here I use the Analytics page wording and read Pin clicks as "opened the closeup". If your screen disagrees, trust Pinterest's current page over me.
Sources: [A1] [A2]. Hands-on verification: pending (V02).
Why these five, and why not the others
You'll end up with a short list and a reason for each, so you can defend it when a tile tempts you to look elsewhere.
My five are impressions, saves, Pin clicks, outbound clicks and the outbound click rate. They sit on one path: a Pin is shown, someone reacts to it, someone opens it, someone leaves Pinterest for your page. The fifth divides outbound clicks by impressions, so periods with different impression counts can be compared.
Why not the rest? Follows and audience figures describe people, not the path to your page, and video metrics apply only to video Pins [A1]. Engagement rate adds saves, Pin clicks and outbound clicks [A1], so it can hide which moved. That is a judgement about getting started, not a claim that those metrics are useless.
Steps 4 and 5 are not Pinterest numbers. I found no Pinterest page saying it counts your visits or orders, so I treat steps 4 and 5 as numbers you hold yourself. The reading sheet below has two lines for them, because a Pinterest number set beside no order count tells half the story.
The five numbers, one question each
You'll end up with what each number can tell you, what it cannot, and the first thing I would check.
1. Impressions: was the Pin shown?
Can tell you: how many times your Pins were on screen in the period. A rise means Pins were shown more times.
Cannot tell you: that anyone looked at the Pin, read it or liked it. Nor can it say why the count changed: Pinterest's page does not explain what drives impressions, so I will not guess at the algorithm.
2. Saves: did someone want to keep it?
Can tell you: the number of times people saved your Pin to a board [A1]. A save is a deliberate act, so I read it as a sign the idea appealed.
Cannot tell you: whether the person will ever visit your page. A save is not a click.
3. Pin clicks: was the Pin opened?
Can tell you: how many times the Pin was opened in closeup, on the Analytics page wording [A1]. In my reading, that is the moment a person can see the full image and its text.
Cannot tell you: that the person went to your page. A Pin can have plenty of Pin clicks and few outbound clicks, and the number cannot tell you whether the closeup answered the question or the link was hard to see.
4. Outbound clicks: did someone act to leave for your page?
Can tell you: the number of times people performed actions that lead to a destination off Pinterest [A1]. This is the closest Pinterest number to "someone headed for my page".
Cannot tell you: that a visit was recorded on your side. Someone may click and close the tab before the page loads, and your stats tool may not count every visit. It also counts only from the date the link was added [A1], so a Pin whose link changed last week has a short history.
First check: put the number beside your own visits for the same dates. If they differ a lot, Make Your Pinterest Click Land on the Right Page covers how to read the gap, so I will not repeat it here.
5. Outbound click rate: out of how many showings?
Can tell you: outbound clicks divided by impressions, in Pinterest's wording [A1]. If impressions double and outbound clicks do not, the rate falls, and that fall is the information.
Cannot tell you: whether the rate is good. A rate is a ratio of two small counts for a small account, so it can swing on a handful of clicks.
First check: the count behind it. A rate of 0.5% from 2,000 impressions rests on 10 clicks, and one click more or fewer moves it by a tenth of itself.
The reading sheet
You'll end up with a sheet you can copy into a notebook. One block per period.
Choose a date range and keep it the same length every time. Pinterest's page says the date range filter covers custom ranges for the past year and a half, plus real-time estimates from the last 48 hours [A1]. I would read finished days and leave the real-time toggle off, because those figures are estimates that might decrease.
Set the content type filter to Organic if you are not running ads. Pinterest describes Organic as Pins created or saved by you, plus Pins created or saved by others that drive people to your claimed website or account [A1]. So the number can include other people's activity, which is a reason to write the filter at the top of the sheet.
PINTEREST READING SHEET
Dates: ____ to ____ (same length
as last period? yes / no)
Filter: Organic / All / other: ____
1 Impressions: ____
2 Saves: ____
3 Pin clicks: ____
4 Outbound clicks: ____
5 Outbound rate: ____ %
(outbound clicks / impressions)
-------------------------------------
FROM MY OWN STATS, SAME DATES
Visits tagged Pinterest: ____
Orders from them: ____
(blank? write: not measured / unknown)
-------------------------------------
CHANGE vs LAST PERIOD
Impressions ___% Saves ___%
Pin clicks ___% Outbound ___%
Changes I made: ____________
-------------------------------------
PATTERN: _________________
DECISION (one): ___________
RE-CHECK ON: ______________
(Tagging a link so your stats show Pinterest visits is explained in Make Your Pinterest Click Land on the Right Page. Without a tag, write not measured.)
Write the number you read. Use 0 only if Pinterest shows zero, not measured if nothing counts it on your side (a possible state for visits), and unknown if a counter exists but shows nothing. Mixing these up turns "I did not set up tracking" into "no one came".
For the change lines, divide the difference by last period's number: (this − last) ÷ last. With small counts, a percentage change can look dramatic. 4 saves against 2 is "+100%" and means two more saves.
If you want a weekly habit, A Fifteen-Minute Weekly Review for a Small Shop gives a fixed slot and rules for small counts.
A worked example: Rin's two months (imaginary)
You'll end up seeing the sheet filled in twice, with every rate recomputed.
Rin is imaginary. She sells printable habit trackers, and her account, her Pins and every number below are invented to show the arithmetic. They are practice figures. They are not benchmarks, not results, and not what any real account gets. I recomputed each figure with Python.
Rin compares September with October, both read as two equal-length ranges with the Organic filter.
RIN (imaginary), organic
Sept Oct
Impressions 8,000 14,000
Saves 120 150
Pin clicks 240 280
Outbound 64 66
Visits (hers) 52 41
Orders 2 1
The rates and changes:
| Sept | Oct | Change | |
|---|---|---|---|
| Impressions | 8,000 | 14,000 | +75.0% |
| Saves | 120 | 150 | +25.0% |
| Pin clicks | 240 | 280 | +16.7% |
| Outbound clicks | 64 | 66 | +3.1% |
| Outbound ÷ impressions | 0.80% | 0.47% | down |
Saves ÷ impressions went from 1.50% to 1.07%, and Pin clicks ÷ impressions from 3.00% to 2.00%. That last figure is my own division. Pinterest's "Pin click rate" definition is worded differently on its page [A1], so I do not claim it matches the tile on your screen.
Impressions rose by 75%. Outbound clicks rose by 2, from 64 to 66. At September's rate, 14,000 impressions would have meant 14,000 × 64 ÷ 8,000 = 112 outbound clicks; Rin got 66. So the extra showings did not turn into extra trips off Pinterest in the way September's did.
Then the other half. Her own visits fell from 52 to 41, so the gap between outbound clicks and visits grew from 12 (64 − 52) to 25 (66 − 41). That is 18.8% of September's outbound clicks and 37.9% of October's. Her orders went from 2 to 1.
What can be said? The pattern is "impressions up, outbound clicks flat, then fewer visits per outbound click". Rin's decision, as my judgement, is to look at the promise and the image of the Pins that gained the most showings, then check the link and page of those Pins using the Pin-to-page worksheet in Make Your Pinterest Click Land on the Right Page.
What cannot be said: that the new Pins are worse, that Pinterest changed who sees them, or that orders fell because of anything Rin did. Orders of 2 and 1 are too small to read: 2 orders in 52 visits is 3.85%, 1 order in 41 visits is 2.44%, and one order more or fewer moves either figure by about two points. I would write "inconclusive" for orders and keep logging.
Sources: none; this section is arithmetic on invented numbers. The script and its output are saved with this guide's notes. Hands-on verification: pending (V03).
One decision per number pattern
You'll end up with a lookup: find your pattern, read the decision, write it on the sheet.
Every decision below is my judgement. None is a Pinterest instruction, none has been tested on a live account, and I have no source saying any of them works. They are the cheapest next checks I can think of for each pattern. Compare equal date ranges, and write "wait" when the counts are tiny.
| Pattern | Decision I would write |
|---|---|
| Impressions up, outbound flat | Look at the Pin's promise and image. |
| Saves up, Pin clicks flat | Check whether the closeup shows what to do next. |
| Pin clicks up, outbound flat | Check the link and the closeup text. |
| Outbound up, visits flat or down | Use the Pin-to-page worksheet. |
| Visits up, orders flat | Audit the page, then run one test. |
| Impressions down | List what changed first. |
| All numbers small | Wait and log again. |
The details for the three patterns I would explain further:
Impressions up, outbound clicks flat. More showings, the same number of departures. Look at the Pin as a stranger scrolling: does the image say in a few words what the click will give? Does the title promise something the page delivers? Change one of them on one Pin, not both on all of them. This is Rin's case above.
Outbound clicks up, visits flat or down. The click leaves Pinterest and nothing arrives on your side. Possible causes are a slow page, a changed link, a missing tag or a stats tool that misses some visits. Make Your Pinterest Click Land on the Right Page has the match worksheet and the log for this. If the link is new, remember Pinterest counts outbound clicks from that date [A1].
Impressions down. Write down what changed first: fewer Pins posted, a shorter date range, the real-time toggle, the filter. Pinterest says real-time estimates might decrease [A1], so a number that drops on the next look may be an estimate settling.
For visits up and orders flat, Audit Your Product Page in 30 Minutes scores the page, and Why Your Shop Isn't Growing shows how to write a one-change test. When all numbers are small, a percentage change tells you very little, so my decision would be to keep the Pins as they are and compare next period. "Wait" is a decision.
Common misreadings
You'll end up with a short checklist of traps, each tied to what Pinterest's pages say or leave unsaid.
- Treating impressions as readers. Pinterest defines impressions as times Pins were on screen [A1]. By that definition, a Pin that scrolls past counts like one someone studies. A different Pinterest page says a paid-ad impression needs one pixel on screen for one continuous second [A3]; it is about ad measurement, and the organic page does not say the same, so I do not apply it.
- Treating engagement as clicks. Engagements include saves, Pin clicks, outbound clicks and more [A1]. A rising engagement count can come entirely from saves, with no extra visits.
- Mixing filters or date ranges. Organic, and Paid and earned, are separate views [A1]. Seven days against 30, or a month with ads against a month without, gives a difference that is only arithmetic.
- Expecting Pin stats and Analytics to match. The Pin stats page says most Pin stats update in real time, so they might look different from Pinterest Analytics [A2]. It also says Pin stats aggregate data for all Pins that share an image and, where applicable, a URL [A2]. A total on one screen can include more Pins than you expect.
- Reading the top Pins list as "best". The page says the list starts sorted by most impressions, and you can change it to engagements, Pin clicks, outbound clicks or saves [A1]. The first Pin on the list answers the question the sort asks, which may not be the question you have.
- Reading a rate without its count. 1 of 20 and 50 of 1,000 are both 5%. Only one of them is a stable number.
Frequently asked questions
What is a good save rate? I do not know. Pinterest's page gives no benchmark and I found no source. Compare with your own earlier periods.
Should I read each Pin or the whole account? Start with account totals to find a pattern, then open the top Pins for the metric that moved. That order is my preference.
Where you are now
You have a reading sheet with five numbers for two equal date ranges, a named pattern and one decision with a re-check date. None of this predicts sales. It tells you which part of the path to look at next.
Your next step
Pick by your pattern. None of these is required.
- Outbound clicks but no visits: Make Your Pinterest Click Land on the Right Page.
- You want a fixed habit: A Fifteen-Minute Weekly Review for a Small Shop.
- Visits but few orders: Audit Your Product Page in 30 Minutes.
- You want the whole picture first: Beginner Sales Funnel in Plain English.
Sources and checks
How we checked. On 26 October 2026 I fetched and read the text of three Pinterest help pages. [A1] is Pinterest Business help, "Review Pinterest Analytics". The same page was also read in the pilot on 25 October 2026, and the metric definitions matched. [A2] is "Review your Pin stats". [A3] is "Pinterest description of methodology", about ad measurement. The saved texts are in this project's raw notes. Nothing here has been tested in a live Pinterest account.
Sources, as checked.
- [A1] Pinterest Business help, "Review Pinterest Analytics", fetched 2026-10-26: the metric definitions, the outbound click date note, the 30-day overview, real-time estimates, filters and top Pins sorting.
- [A2] "Review your Pin stats", fetched 2026-10-26: the alternative Pin click wording, aggregation of Pins sharing an image and URL, and that Pin stats may differ from Analytics.
- [A3] "Pinterest description of methodology", fetched 2026-10-26: ad impression counting only.
Method and opinion. The five numbers, the sheet, the pattern table and every decision are my judgement. Rin and her numbers are invented; her rates were recomputed in Python.
Unverified. What drives impressions, any benchmark or good rate, how the two Pin click wordings relate, whether the Pin click rate tile equals Pin clicks divided by impressions, how organic impressions are counted, and any change to Pinterest's pages after the fetch date. A web search limited to help.pinterest.com for further definition pages returned the same pages plus ad reporting pages, which I did not read in full.
Checks still to run. V01 to V03 in the front matter. Until real sellers have used the sheet, this guide stays in review.
What this guide does not claim. That any rate is normal, that any pattern means what the decision assumes, that Rin existed, or that reading these numbers will raise traffic or sales.