Definition
Reach is the deduplicated count of people exposed to an ad, ad set, campaign or account at least once during a period. The deduplication is what distinguishes it from impressions: one person who sees the same ad several times contributes several impressions but one unit of reach. What counts as a person depends on the platform's identity system. A platform with logged-in users can count accounts and, with some modelling, people; the open web mostly counts browsers or devices, and a person with a phone and a laptop is two units unless the measurement system can link them. For that reason most platforms label reach as an estimate, and reach figures for the same campaign from two measurement sources rarely agree.
How it works
The platform records each impression against an identifier, collapses identifiers it believes belong to the same person using its identity graph, and counts the distinct results within the reporting window. Where the identifier is missing or the platform's confidence is low, it models the count from a sample or a panel. Because the window matters, reach for a week is not the sum of seven daily figures: the same person appears on several days and is counted once in the weekly number. The same logic stops reach being added across ad sets, campaigns or platforms, where overlap is unknown; the cross-platform reach and frequency guide covers the identity methods and the overlap problem.
| Identity basis | Typical environment | What "one person" means | Reliability |
|---|---|---|---|
| Logged-in account | Social platforms, video platforms | An account, modelled toward a person | High within the platform |
| Cookie or device identifier | Open-web display, app inventory | A browser or a device | Lower; falls with cookie and identifier restrictions |
| Panel or census model | Cross-platform measurement vendors | A statistically estimated person | Depends on panel size and modelling |
Reach and FrequencyAverage times a person saw an ad. are linked by impressions: impressions equal reach multiplied by average frequency, so for a fixed budget, widening reach lowers frequency and the reverse. Planning tools and the media planning framework set the intended balance before the campaign runs, usually with an effective reach target, the number of people exposed at least a chosen number of times rather than at least once.
In practice
Reach is the primary metric for awareness campaigns, where the objective is to expose as many of the right people as possible, and a diagnostic metric for everything else, where it explains why frequency rises or why results stall. The practical reading is always reach against the size of the audience that could have been reached: a campaign that has covered most of its target audience has nowhere to grow except by changing the audience, and further spend buys frequency instead. The reach and frequency calculator turns any pair of these figures into the third.
Reach that stopped growing
A video campaign spends £40,000 over four weeks against an audience the platform sizes at 1,500,000 people. Weekly reach reads 520,000, then 790,000, then 910,000, then 960,000, while weekly impressions stay near 1,800,000. Cumulative reach is 64% of the audience, and in the fourth week only 50,000 new people were reached for £10,000, a cost of £0.20 per new person against £0.019 in the first week. Monthly frequency has climbed to 7.5. The planner judges that effective reach at three or more exposures, which stands at 71% of those reached, is high enough, and moves the final week's budget to a second audience rather than buying more repeat exposures.
What it is not
Reach is not impressions, and the gap between the two is frequency. It is not "potential reach" or "audience size", the planning estimate of how many people match a targeting definition before anything is served. It is not the "users" metric in web analytics, which counts people who visited a site rather than people who saw an ad, and which uses a different identity system. It is not awareness: a person counted as reached may not have noticed the ad, which is what viewability standards and brand lift studies attempt to measure separately. It is not additive, so a report that sums reach across campaigns or channels overstates the number of distinct people, sometimes by a large margin. And it is not comparable across platforms, because each one defines a person and an exposure in its own way.
Related terms
FrequencyAverage times a person saw an ad. is reach's twin: the two share the impression count between them, and most reach problems are described more precisely as frequency problems. CPMCost per 1,000 impressions. prices the impressions, but the cost that matters for an awareness objective is cost per person reached, or cost per person effectively reached, which the platform's reach report or the calculator derives. Share of VoiceYour share of market visibility. is the strategic cousin, the share of a category's advertising exposure that belongs to one brand, for which reach within the category's audience is the usual input. The core metrics guide places reach among the exposure metrics and shows how it chains into response and outcome metrics, and the cross-platform guide covers unique reach reports and what they assume. Impressions, the raw count behind both reach and frequency, is the metric most often mistaken for reach in reporting, and keeping the three on the same line of a report is the simplest guard against that confusion.
Common mistakes
- Adding reach across campaigns or platforms. Overlap is unknown and usually large; use a deduplicated report or a model.
- Reporting reach without the period. A monthly figure and a weekly figure describe different things.
- Treating the platform's estimate as a count. It is modelled, and its accuracy varies by environment.
- Confusing audience size with reach. The first is who could be reached, the second who was.
- Judging an awareness campaign on reach alone. Effective reach and frequency distribution say whether the people reached saw enough to notice.
- Comparing reach between platforms as if it were one metric. A logged-in platform and open-web display define a person differently, so the larger number is not necessarily the larger audience.
- Letting reach grow by widening targeting silently. Automated placements and expansion settings raise reach by adding people outside the intended audience, which the top-line figure does not reveal.
