Reach is the number of unique people or households exposed to an ad or campaign at least once during a given period. Each person is counted once, however many times they saw the ad. Reach is often expressed as a percentage of the target population and is reported together with frequency, the average number of exposures per person reached.
Reach, impressions and frequency
- Reach: number of unique people exposed.
- Impressions: total number of times the ad was displayed, including repeats.
- Frequency: average number of exposures per person reached.
The three are linked by an identity: Impressions = Reach × Average frequency.
Example: a campaign delivers 1 200 000 impressions to 400 000 unique people in a target population of 2 000 000. Reach is 400 000, reach % is 20% and average frequency is 1 200 000 ÷ 400 000 = 3.0. In traditional media planning, the same campaign would be described as 20 × 3.0 = 60 gross rating points (GRP = Reach % × Average frequency).
Types of reach
Gross (summed) reach: the sum of the reach of each channel or campaign without removing people reached by more than one. It overstates the audience whenever channels overlap.
Net (deduplicated) reach: the number of unique people reached across channels. This is the number most useful for planning, and the hardest to measure.
Incremental reach: the additional unique people reached by adding a channel or extending a campaign. Note that this is a reach concept, not incrementality in the causal sense: reaching new people does not by itself prove additional sales.
Effective reach: the number of people exposed at least a chosen minimum number of times, the "effective frequency". The common 3+ threshold traces back to Krugman's argument that three exposures may be enough (Krugman, 1972) and to the Association of National Advertisers review edited by Naples (1979). Later work, such as Tellis (1997), argues that no single number applies universally: the effective level depends on factors such as brand familiarity, message complexity and novelty. Treat 3+ as a planning convention, not a law.
How reach is measured
TV, radio and print: reach is estimated from audience panels and surveys, then expressed as ratings and GRPs.
Digital platforms: platforms report unique users or accounts, but these numbers are usually modeled. Google, for example, notes that its reach and frequency metrics rely on statistical models and require minimum impression and reach thresholds (Google Ads Help, Measuring reach and frequency).
Across platforms: each platform deduplicates only within its own ecosystem. Adding up reach from several platforms gives gross reach, not net reach. Cross-device use and privacy restrictions on identifiers make deduplication harder still.
Reach curves and diminishing returns
Reach does not grow linearly with spend. The first impressions of a campaign mostly reach new people; as the campaign continues, a growing share of impressions goes to people already reached, so frequency rises and new reach slows. Plotting cumulative reach against impressions or budget gives a reach curve that flattens as it approaches the reachable audience. The point where the curve flattens is where extra budget mainly buys frequency.
Reach and frequency in marketing mix modeling
Marketing mix modeling (MMM) estimates how incremental sales respond to media activity. Most MMMs use spend or impressions as the media input and apply a saturation function to capture diminishing returns (Jin et al., 2017). The flattening reach curve is one reason those response curves bend: as spend grows, more impressions go to people who have already seen the ad several times.
When reach and frequency data are available, they can be modeled separately. Google researchers proposed an extension of geo-level Bayesian hierarchical MMM that uses reach and frequency as inputs for channels where that data exists, and reported more accurate estimates and the ability to recommend an optimal frequency (Wurm, Liu et al., 2023). Google's open-source Meridian MMM implements this: it treats the effect of reach as linear and applies a Hill function to average frequency, which allows an S-shaped response to frequency (Google Meridian documentation). In practice this separates two questions a spend-based curve mixes together: should we reach more people, or reach the same people more often?
Reach is still an exposure metric. To know whether broader reach produced additional sales, the reach data need to be connected to sales through MMM or tested with experiments such as geo tests.
Key takeaways
- Reach counts unique people; impressions = reach × average frequency.
- Summed platform reach is gross reach; deduplicated net reach is hard to measure across platforms.
- The 3+ effective reach threshold is a convention from Krugman (1972) and Naples (1979), not a universal rule.
Related terms
- Frequency
- Impressions
- Viewable impressions
- Impression share
- Cost per mille (CPM)
- Incrementality
- Adstock
Further reading on Sellforte
- Advertising response curves: what are they and why do you need them?
- How to measure upper-funnel marketing in ecommerce
- How to measure offline media with marketing mix modeling
- OTT and CTV measurement
References
- Krugman, H. E. (1972). Why three exposures may be enough. Journal of Advertising Research, 12(6), 11–14. https://cir.nii.ac.jp/crid/1574231874476450560
- Naples, M. J. (1979). Effective frequency: The relationship between frequency and advertising effectiveness. Association of National Advertisers. https://worldradiohistory.com/Bookshelf/Business/Effective-Frequency-ANA-Naples-1979.pdf
- Tellis, G. J. (1997). Effective frequency: One exposure or three factors? Journal of Advertising Research. https://www.warc.com/content/article/jar/effective-frequency-one-exposure-or-three-factors/en-gb/9066
- Google. (n.d.). Measuring reach and frequency. Google Ads Help. Retrieved 2026-10-30, from https://support.google.com/google-ads/answer/2472714
- Jin, Y., Wang, Y., Sun, Y., Chan, D., & Koehler, J. (2017). Bayesian methods for media mix modeling with carryover and shape effects. Google Research. https://research.google/pubs/bayesian-methods-for-media-mix-modeling-with-carryover-and-shape-effects/
- Wurm, M., Liu, Y., et al. (2023). Bayesian hierarchical media mix model incorporating reach and frequency data. Google Research. https://research.google/pubs/bayesian-hierarchical-media-mix-model-incorporating-reach-and-frequency-data/
- Google. (n.d.). Reach and frequency. Meridian documentation. Retrieved 2026-10-30, from https://developers.google.com/meridian/docs/basics/reach-frequency
Authors
Lauri Potka is the Chief Operating Officer at Sellforte, with over 15 years of experience in Marketing Mix Modeling, marketing measurement, and media spend optimization. Before joining Sellforte, he worked as a management consultant at the Boston Consulting Group, advising some of the world’s largest advertisers on data-driven marketing optimization. Follow Lauri in LinkedIn, where he is one of the leading voices in MMM and marketing measurement.
