Quick answer
Reach is the number or percentage of distinct people or households exposed to advertising at least once during a defined period. Average frequency is total impressions divided by reached people. Gross Rating Points, or GRPs, summarize campaign weight and can be calculated as percentage reach multiplied by average frequency. A plan reaching 40% of its defined audience 2.5 times on average delivers 100 GRPs. Target Rating Points apply the same logic to a specified target. GRPs count duplicated exposure, not unique people, so 100 GRPs can come from many different reach-frequency patterns. Inspect the full frequency distribution, measurement source, time window and delivered quality. There is no universal rule that exactly three exposures work for every objective, category, creative or medium.
What do reach, frequency and GRPs mean?
Reach counts distinct members of a defined audience exposed at least once in a defined period. It can be reported as a number or percentage. Frequency describes how many exposures reached people receive. GRPs express the total duplicated rating weight delivered by a schedule.
These are media-delivery concepts, not direct measures of attention, memory or persuasion. An opportunity to see or hear may meet the currency definition without being noticed. Quality and creative response require additional evidence.
The audience unit also matters. Household reach can include several possible viewers, while people-based reach attempts to estimate individuals. Device reach counts devices unless identity methods infer people. Never switch units inside one calculation or describe device cookies as confirmed human beings.
Every report needs a population, geography, period, medium and source. Reach among adults in one city over a week is not comparable with household reach nationally over a month without adjustment.
GRPs became a common currency for broadcast planning
Ratings systems gave broadcast advertisers a standardized way to estimate audiences for programs and commercial schedules. Planners summed rating points to express campaign weight and used modeled duplication to estimate reach and average frequency.
Digital media added impression-level logs and platform-defined reach, but identity fragmentation complicated unique-person counting. One person can use several devices or appear in several systems, while shared devices can represent several people.
Cross-media initiatives now seek privacy-conscious deduplication across television and digital exposures. Until methods and coverage are aligned, planners should not add self-reported platform reach as if every person were unique.
Deduplication requires an identity or statistical bridge. Panels, clean rooms, device graphs and virtual-person models each make trade-offs between coverage, privacy, latency and error. The planner should know which method produced the number and whether it covers every channel in the reported total.
Use the three formulas with aligned units
Reach percentage equals unique people reached divided by the defined target population, multiplied by 100. Average frequency equals total counted impressions divided by unique people reached. GRPs equal percentage reach multiplied by average frequency, or the sum of rating points across insertions.
If 400,000 people in a one-million-person target receive one million impressions, reach is 40%, average frequency is 2.5 and weight is 100 GRPs. If the calculation uses a narrower target population, the result is a Target Rating Point, commonly called a TRP.
Keep percentages and points clear. Forty percent multiplied by 2.5 gives 100 rating points, not 100% unique reach. Report rounding and whether impressions include co-viewing, modeled people, viewability or other filters.
Inspect frequency distribution, not only the mean
Average frequency compresses an uneven pattern. A campaign averaging three exposures may leave many people at one exposure while a smaller group receives ten. Examine cumulative reach at 1+, 2+, 3+ and higher levels, plus the share above a waste or annoyance threshold.
Two plans can both deliver 100 GRPs. One might reach 50% twice; another might reach 25% four times. The preferable shape depends on whether the task needs broad memory coverage, repeated instruction, a narrow eligible group or a time-sensitive response.
Look at spacing as well as count. Three exposures within one minute do not create the same opportunity to retrieve a brand as three exposures across relevant days. Most standard frequency reports omit spacing, so use schedule and log-level diagnostics where lawful and available.
Use incremental reach curves to see when another channel adds new people rather than more duplication. Because cross-media curves are often modeled, include confidence and compare forecast with delivered results.
Show the cumulative curve beside the frequency histogram. The curve reveals how quickly new people are added as spend grows, while the histogram reveals where repeated impressions accumulate. Together they expose plans that appear to extend reach at the margin but still place most additional delivery on already heavy users. Review both before reallocating budget.
There is no universal effective-frequency threshold
The common three-exposure rule is frequently repeated as if it were a law. Gerard Tellis's review argues that effective frequency depends on multiple factors rather than one fixed number. Creative familiarity, message complexity, brand knowledge, category involvement, competition, medium and timing all matter.
One exposure can generate an effect, especially for a familiar brand or clear offer. Additional exposures may build memory, clarify a message or enable action, then reach diminishing returns or wear-out. The response curve can differ across people.
Set a frequency hypothesis and test it. Compare outcomes by exposure bands cautiously because people with more exposures may differ from those with fewer. Randomized frequency experiments or platform tools can improve inference where available.
Specify the outcome window before testing. A frequency pattern associated with immediate site visits may differ from one that builds branded memory over weeks. Report confidence intervals and sample size by exposure band because the highest-frequency groups can be small, unusual and especially vulnerable to selection bias.
Balance reach and repetition against the objective
For broad brand building, prioritize relevant unduplicated reach while maintaining enough continuity for memory. For a complex new behavior, some repeated explanation may be valuable. For a small eligible population, higher frequency can be reasonable if caps and consent protect experience.
Forecast several scenarios at the same budget or GRP weight. Compare reach curve, frequency distribution, timing, channel overlap, creative rotation and cost. A cheap high-frequency plan may simply saturate easy-to-buy users.
Separate frequency for acquisition and current customers. Suppression can reduce obvious waste, but identity gaps and shared devices make it imperfect. Where suppression is unavailable, choose contexts and creative that remain respectful even when a loyal user sees the message again.
Plan creative alongside exposure. Repeating identical assets is different from repeating a recognizable idea through fresh executions. Rotation can reduce wear-out while retaining distinctive cues and message continuity.
Hypothetical example: a local cinema membership
A fictional independent cinema wants more local residents to consider a monthly membership. One modeled plan delivers 100 GRPs through 50% reach at average frequency two. Another also delivers 100 GRPs through 25% reach at average frequency four, concentrated among current frequent filmgoers.
Because the objective includes new occasional visitors, the first scenario is the stronger starting hypothesis. The planner adds local outdoor and broad video to extend reach, then uses search and site follow-up for people actively exploring schedules. Existing members are suppressed where possible.
The example is entirely hypothetical and reports no effect. The campaign compares forecast and delivered reach, 1+ and 3+ distributions, branded memory, qualified visits and incremental membership, while monitoring excessive exposure and geographic spill.
Reconcile forecasts, delivery and quality
Pre-campaign estimates come from panels, planning tools, historical data and modeled duplication. After launch, compare delivered impressions, reach, frequency and cost against forecast using consistent definitions. Explain differences caused by auction, inventory, audience availability or identity resolution.
Layer quality: on-target delivery, viewability, invalid traffic, placement, duration, audibility and attention proxies where appropriate. A viewable impression is not proof of attention, but unviewable or invalid delivery cannot perform the intended communication job.
Reconcile denominators before comparing vendors. One system may report reach among logged-in users, another among a modeled population and another among panel households. Place all methodology beside the number and avoid ranking plans on point estimates whose confidence ranges overlap materially.
Connect media metrics to brand and behavior outcomes using experiments or other causal methods. Avoid claiming that a high GRP total caused sales merely because both occurred in the same period.
Limitations and common misuse
Reach estimates depend on identity and measurement coverage. Panels have sampling error, digital logs can duplicate people and co-viewing is modeled. Different vendors may produce different answers for the same campaign. Report method and uncertainty.
GRPs treat rating points as additive even when contexts, formats and creative quality differ. One hundred points of short muted feed impressions are not automatically equivalent to 100 points of television exposure. A common number does not erase exposure conditions.
Optimization can also hide exclusion. Frequency caps may work only inside one platform, while the same person receives ads elsewhere. Cross-channel governance and suppression remain imperfect, so monitor complaints and repeated-user samples.
GRPs tell you how much duplicated audience weight was delivered. They do not tell you whether the right people noticed, remembered or acted.
Reach, frequency and GRP checklist
Use this checklist before comparing plans or presenting delivered audience metrics.
- Target population and geography are defined
- Measurement period is explicit
- Reach is unique within the stated system
- Impression definition is documented
- Frequency formula uses aligned data
- GRPs use percentage reach, not a decimal mistake
- TRP target is named where relevant
- 1+ through higher exposure bands are shown
- Over-frequency threshold is monitored
- Cross-media duplication is modeled transparently
- Forecast and delivered metrics are reconciled
- Quality and invalid traffic are reported
- Creative rotation is planned
- Effective-frequency claim is tested
- Incremental outcome is measured separately
Frequently asked questions
What is advertising reach?
Reach is the number or percentage of distinct people or households exposed at least once during a defined period.
How is average frequency calculated?
Divide total counted impressions by the number of unique people reached within the same population, medium scope and period.
How are GRPs calculated?
Multiply reach expressed as a percentage by average frequency. Forty percent reach at 2.5 frequency equals 100 GRPs.
What is the difference between GRP and TRP?
GRP often refers to the overall measured population, while TRP applies the same rating-point logic to a defined target audience.
Is three exposures the ideal frequency?
No. Effective frequency varies with objective, creative, familiarity, category, medium, competition and timing, and should be tested rather than assumed.
Sources and further reading
- Google Ads Help: Cross-Media Reach Planner Metrics ↗First-party definitions of reach, frequency, GRPs and TRPs used in cross-media planning
- World Federation of Advertisers: Cross-Media Measurement Framework ↗Advertiser-led framework for deduplicated reach and frequency across media
- SSRN: Effective Frequency, One Exposure or Three Factors? ↗Gerard Tellis's evidence-based challenge to a universal exposure threshold
- Media Rating Council: Cross-Media Audience Measurement Standards ↗Independent standards for audience, exposure quality, invalid traffic and cross-media video measurement