Reach
By the AdFlint research team · Last reviewed July 2026
The number of unique people who saw an ad at least once during a specified period, deduplicated across all of its impressions.
Reach and impressions coincide only at a frequency of one. It is estimated and tied to accounts or devices rather than humans, so shared devices and duplicate accounts distort it. It also cannot be summed across date ranges or campaigns without counting the same people twice. The common misreading is adding weekly figures together to claim a monthly total.
Key takeaways
- Reach and impressions are only equal at frequency 1; once people see an ad more than once they diverge, and reach is always the smaller or equal number.
- Never sum reach across weeks, months, or campaigns to get a total - overlapping people get double-counted, so only a platform's own deduplicated report at that grain is trustworthy.
- Conversion-optimized campaigns tend to cap reach growth even with unspent budget, because the algorithm favors repeat exposure to likely converters over finding new people.
- A plateauing reach curve with rising frequency and rising CPM signals a saturated audience - widen targeting or refresh creative rather than raising bids.
In practice.
Reach counts unique people, or more precisely unique accounts or devices Google and Meta believe correspond to unique people, who saw at least one impression during a set window. The distinction from impressions matters mechanically: an impression fires every time an ad renders, so one person scrolling past your ad five times in a day generates five impressions but only one unit of reach. The two numbers converge only at a frequency of exactly one - the moment your audience starts seeing the ad more than once, reach and impressions diverge, and the ratio between them is Frequency (impressions divided by reach).
Both platforms estimate reach rather than count it with certainty, and that estimation is where most of the confusion starts. Meta and Google deduplicate based on logged-in accounts, device identifiers, and modeled matching across devices, not verified human identity - a shared family tablet, a person logged into two ad accounts, or someone who cleared cookies and got re-counted as a new person all introduce distortion in either direction. Reduced device-level signal from iOS privacy restrictions has made cross-device reach modeling noisier than it used to be, so reach figures should be read as a directional estimate, not a precise headcount, especially at the edges of small audiences or short flights.
The setting that shapes reach most directly is your campaign's frequency cap, where available, and your bid strategy's optimization goal. A campaign optimizing for reach or awareness will spread delivery across as many unique people as the budget allows, actively working against high frequency. A campaign optimizing for conversions will do the opposite by default - it happily shows the same converting-likely person the ad repeatedly if that is where the algorithm finds the best return, which caps reach growth even with budget left to spend. This is why a conversion campaign and an awareness campaign with identical budgets in the same account can post very different reach numbers: the objective, not the spend, is doing most of the work.
The number one practitioner error is treating reach as additive across time periods or campaign structures. Adding reach from week one and week two of a flight does not produce a monthly reach total, because plenty of the same people were counted in both weeks and would be double-counted in the sum - platforms report monthly reach only when you pull the report at that grain, using their own deduplication across the full period, not by you doing the arithmetic. The same problem applies across campaigns or ad sets: summing per-campaign reach numbers overstates true unique reach whenever campaigns share any audience overlap, which broad and interest-based Meta targeting almost always does to some degree.
In reporting, reach is most useful paired with Frequency to judge whether a flight is still finding new people or has started recycling the same pool at rising cost - a plateauing reach curve alongside climbing frequency and climbing CPM is the classic signature of a saturated audience, distinct from Impressions or Impression Share, which describe volume and competitiveness rather than uniqueness. Reach alone says nothing about response; pairing it with CTR, Engagement Rate, or Conversion Rate is what tells you whether the people you reached actually did anything, and reach should never be the only metric used to judge whether a campaign is working.
Why weekly reach totals don't add up to a monthly number
Suppose a local retailer's Meta campaign reports reach of 40,000 in week one and 38,000 in week two, targeting the same fixed audience of roughly 90,000 eligible people in the metro area. Adding those figures naively suggests 78,000 unique people were reached over the month.
But when the retailer pulls the campaign's actual 30-day reach report, Meta shows 52,000 unique people, not 78,000 - because a large share of the people counted in week one were shown the ad again in week two and counted again in that week's total, not as new people. With frequency climbing from 1.3 in week one to 2.1 by week two, the campaign was increasingly re-showing the same pool rather than finding fresh viewers, which is exactly why the deduplicated 30-day number comes in well below the sum of the weekly figures.
Reach compared with
The settings this gets confused with, and how to tell them apart.
Common questions.
Why does my reach number look smaller than my impressions number?
That's expected whenever frequency is above 1 - impressions count every ad view, reach counts unique viewers, so reach will always be equal to or less than impressions. The gap between them grows as the same people see the ad more times.
Can I add up daily or weekly reach to get a campaign total?
No, because the same people are likely counted in multiple days or weeks, so summing overstates true unique reach. Pull the report at the actual date range you want (for example, the full 30-day window) and let the platform's own deduplication produce that number.
Why is my reach stuck even though I still have budget left to spend?
This usually means your campaign is optimizing for conversions or another action rather than reach, so the algorithm is choosing to show the ad again to people likely to convert instead of finding new viewers. Switching objectives toward reach or awareness, or widening the audience, will unstick it if finding new people is actually the goal.
How accurate is the reach number platforms report?
It's a modeled estimate based on account and device signals, not a verified human count, and accuracy has gotten noisier since iOS privacy changes reduced cross-device signal. Treat it as directional, especially for small audiences or short flights, rather than as an exact headcount.
You should not need to know this to advertise.
AdFlint handles the settings for you, inside the Google and Meta accounts you already own.
Try AdFlint free