To read rolling reach, pull Meta reach for a trailing window that ends today, pull it again for the same length window that ended one week earlier, and subtract. Do that for 7, 14, 30 and 90 day windows. The short windows tell you how your current ads and budget are doing right now; the long windows tell you whether the pool of people you reach is actually growing.
A note on where this comes from: we did not invent these reach reports. The methods and formulas here are pulled together from Meta's own tools and documentation, reports we have seen across the industry, and other practitioners' work. We are sharing how we use them, not claiming to be the originator.
Definition
Rolling reach is the number of unique people your ads reached over a moving window, like the last 7, 14, 30 or 90 days, pulled on the same weekday every week.
If you are new to the idea of net new reach, start with what net new reach is and how to measure it. This article goes one level deeper: what the subtraction is really measuring inside each window, how to pull the numbers each week, and how to read four windows together.
What is a rolling reach window?
A rolling reach window is a trailing date range of fixed length, such as the last 30 days, that slides forward each time you report. Because it always covers the same number of days, you can compare it fairly from one week to the next.
Here is the problem it solves. If you compare this week's reach to last week's reach, you learn whether you showed ads to more people this week. You do not learn whether those people were new. A rolling window fixes that, because Meta counts each person once across the whole range you pull. Reach for "the last 90 days" is the number of unique people who saw at least one of your ads in those 90 days, no matter how many times they saw them.
Why use four windows instead of one?
Each window length changes what "new" means, so each one answers a different question. One window alone can mislead you; four together tell a story.
- 7 day. A person counts as new if they were not reached in..: the prior week. The question it answers: Is weekly delivery growing or shrinking?
- 14 day. A person counts as new if they were not reached in..: the prior two weeks. The question it answers: Are this week's ads reaching people the last fortnight missed?
- 30 day. A person counts as new if they were not reached in..: the prior month. The question it answers: Is fresh creative opening new pockets of people?
- 90 day. A person counts as new if they were not reached in..: the prior quarter. The question it answers: Is the account still adding to the pool of people it can sell to?
We use these four for weekly work because they line up with the date presets most people already use in Ads Manager. For monthly reporting we slide longer windows by a month instead, up to a full year, as described in the net new reach article. The logic is identical. Only the length and the step change.
What does the subtraction actually measure?
The difference between two rolling pulls equals the people who entered the window minus the people who aged out of it. That is the single most important thing to understand about rolling reach, and it is why the long windows are more honest than the short ones.
Take the 90 day window. The window ending this Sunday and the window ending last Sunday share 83 days. They differ by one week on each end:
- This week's window adds the newest 7 days.
- It drops the oldest 7 days from last week's window.
So the subtraction works out to:
Rolling change = people reached in the newest week who were not reached in the 83 shared days, minus people reached only in the dropped week and never in the 83 shared days.
When that number is positive, more new people came in than old people fell out. Your 90 day audience grew. When it is near zero, the pool is holding steady. When it is negative, the pool shrank.
The shortest window is a special case. A 7 day window ending this Sunday and one ending last Sunday share no days at all, so the 7 day subtraction is simply this week's reach minus last week's reach. It is a pace reading, and it is driven mostly by budget and delivery. Treat it as context, not as proof of growth.
As windows get longer, they overlap more, and the subtraction gets closer to a clean count of "people new to us over this stretch." That is why we trust the 90 day read more than the 7 day read when the two disagree.
How do you pull rolling reach in Meta Ads Manager?
You need eight numbers each week: two pulls for each of the four windows. Pull them at the account level, across every campaign, so restructures and renamed campaigns do not break your comparisons.
- Open Ads Manager for the ad account and go to the Campaigns tab without selecting any campaign.
- Make sure the view includes every campaign that spent in the period, not only active ones. Paused campaigns that delivered still count.
- Under Columns, customize to show Reach, Impressions, Frequency and Amount spent.
- Set a custom date range for the first pull, for example the 90 days ending this Sunday.
- Read the Reach figure in the totals row at the bottom of the table. That total is deduplicated across the campaigns shown. Do not add up the reach of individual campaigns, because one person reached by two campaigns would be counted twice.
- Change the date range to the 90 days ending last Sunday and read the total again.
- Repeat for the 30, 14 and 7 day windows.
- Record spend for the newest week too. You will need it for cost per 1,000 net new reach.
Always pull on the same day of the week and end every window on the same day. Small inconsistencies in the dates add noise that looks like a signal.
One practical tip: do not use the time breakdown to get weekly reach and then add the weeks up. Each row of a breakdown shows reach for that row alone, and the same person appears in several rows. Only a single pull across the full range gives you deduplicated reach.
What does a worked example look like?
Here is a full weekly update with invented round numbers. None of these come from a real account.
- 7 day. Pull A (ending this Sunday): 100,000. Pull B (ending last Sunday): 80,000. Rolling change (A minus B): 20,000
- 14 day. Pull A (ending this Sunday): 170,000. Pull B (ending last Sunday): 150,000. Rolling change (A minus B): 20,000
- 30 day. Pull A (ending this Sunday): 320,000. Pull B (ending last Sunday): 305,000. Rolling change (A minus B): 15,000
- 90 day. Pull A (ending this Sunday): 700,000. Pull B (ending last Sunday): 698,000. Rolling change (A minus B): 2,000
Show the lines the same way in your spreadsheet:
- 7 day: 100,000 minus 80,000 = 20,000
- 14 day: 170,000 minus 150,000 = 20,000
- 30 day: 320,000 minus 305,000 = 15,000
- 90 day: 700,000 minus 698,000 = 2,000
Say the newest week's spend was $10,000. You can now add cost per 1,000 net new reach for each window using that one week of spend:
- 30 day: $10,000 ÷ 15,000 × 1,000 = $667
- 90 day: $10,000 ÷ 2,000 × 1,000 = $5,000
How to read it: weekly reach jumped, and the short and medium windows grew with it. But the 90 day window barely moved. The account is reaching people it has not seen in a week or a month, yet almost all of them were reached sometime in the last quarter. More money bought more delivery to a familiar crowd, not a bigger audience. The cost per 1,000 at the 90 day horizon makes that obvious. For more on that metric, see cost per 1,000 net new reach.
How do you read the four windows together?
Read the windows as a set and look for the pattern, not the single number. These are the patterns we look for first.
All four windows rising together. The account is adding people at every time horizon. This is the healthy pattern, and if your unit economics hold up, it supports more spend.
Short windows up, 90 day flat. Fresh creative or a budget increase is reaching people who have not seen you lately, but they are mostly people you reached within the quarter. Common when an account scales on a small audience base. Not a crisis, but not growth either.
Short windows falling, 90 day steady. The near term is saturating. Weekly delivery is going back to people who just saw you. This is usually the moment to look at exclusions (for example, people who engaged with your page or ads recently) and at whether new creative is landing.
Every window jumps the week after a new ad launches. That ad is reaching people your other ads were not. This is one of the strongest signals in an account. We look hard at scaling an ad like that, even before its purchase numbers mature. The creative reach report is how you confirm which ad did it.
Every window drops at once, along with spend or delivery. This is almost never the audience running out. It is a budget cut, paused campaigns or a structural change. Real saturation looks slower: reach holds steady while the long windows fade week after week.
A negative number in a long window. Check your pulls first, since a wrong end date is the usual cause. If the numbers are right, more people aged out of the window than came in. The pool you are reaching is shrinking, which deserves attention fast.
How does frequency help you read the windows?
Frequency is the average number of times each person saw your ads in a period. It tells you whether flat rolling reach is a saturation problem or a constraint problem.
If the long windows are flat and ad level daily frequency is climbing, the account is showing the same people your ads more often. That points to saturation or weak exclusions. If the long windows are flat and frequency is also flat, the account is probably held back by budget or by creative that does not appeal beyond your existing fans. The fix for each is different, which is why we never read rolling reach without frequency next to it. Our post on daily frequency as a funnel signal explains how to read frequency at the ad level.
Key takeaways
- A rolling reach window is a fixed length trailing range, and the change between two pulls equals people who entered the window minus people who aged out.
- The 7 day change is just this week's reach versus last week's, so it measures pace. The 90 day change is the best weekly read on real audience growth.
- Pull at the account level, from the totals row, on the same weekday, and never add reach across rows.
- Short windows up while the 90 day window stays flat means more delivery to familiar people, not a bigger audience.
- Use one step of spend to price each change, and read frequency alongside reach to tell saturation from constraint.
FAQ
Can rolling reach be negative? Yes, in the longer windows. It means more people dropped out of the window than entered it. Check your date ranges first, and if they are right, treat it as a sign the pool of people you reach is shrinking.
Why not just use Meta's reach column for the last 7 days? That number tells you how many people you reached this week. It does not tell you how many of them were new to you. Comparing overlapping windows is what separates new people from repeats.
Should I pull weekly or monthly? Weekly with 7, 14, 30 and 90 day windows when you are actively managing an account or fixing a problem. Monthly with longer windows for client reports and the big picture. The method is the same either way.
Does rolling reach work at the campaign level? Yes. Meta deduplicates reach inside any single campaign pull, so you can run the same windows on one campaign, such as a Scaling Campaign. Just never add reach across campaigns. See net new reach for a Scaling Campaign.
How many weeks of data do I need before trusting the pattern? More than one. A single week can move on a sale or a delivery glitch. When a pattern holds across several consecutive weeks, it is usually real.
What is rolling reach?Reach over a moving window that moves forward each time you pull it. Comparing windows shows whether your ads keep finding new people.
Want us to run the windows on your account?
If you want to see whether your Meta spend is still growing your audience, we offer a free Meta account audit. It includes a full account audit plus net new reach and cumulative reach reports, so you can see exactly where new people are coming from. Request yours here.
Keep reading
More on reach reports:
Net New Reach Percentage: What Share of Your Meta Reach Is Actually New?
Which Meta Ads Find New People? How to Build a Creative Reach Report
See every reach report article
Ready for your free Meta audit?
We will review your account and run net new reach and cumulative reach reports, free.
Check what you learned
Tap an answer to see if you got it right.
1. Which rolling window gives the best weekly read on real audience growth?
2. The 7 day window goes up but the 90 day window stays flat. What does that mean?
3. How should you pull rolling reach each week?
Ready for the full test? Take the Reach Reports exam and get certified. Want this run on your account? Get my free audit.
Module 4: Reach Reports · article 2 of 15
Next: Net New Reach Percentage: What Share of Your Meta Reach Is Actually New?
Previous: What Is Net New Reach on Meta and How Do You Measure It?
About the author
Jake Mehani, founder of Creative Analytics
Jake has spent 15 years running Meta and paid social ads for consumer brands. Meta published his RBX Active Advantage+ case study, he taught the DMAnc Meta and TikTok ads masterclass, and he spoke at Motion's 2024 Creative Strategy Summit. His senior team runs media buying, creative and measurement like an in house team.