How reach and frequency are calculated for PSA campaigns

Reach is the number of different people, or households, who saw or heard your public service announcement at least once during a campaign. Frequency is how many times, on average, each of them was exposed. Together they answer the two questions a board asks about any PSA campaign: how many people got the message, and how often did they get it?

This guide explains what each figure means, the formulas that tie reach and frequency to GRP and impressions, what moves them up or down, and how to use them when you plan and report a campaign. For broadcast PSAs both figures are estimates, so it also covers how to present them.

What is reach?

Reach is the unduplicated audience of a schedule. Each person or household counts once, however many airings they saw. Planners also call it net reach or 1+ reach, meaning exposed one or more times.

Reach is usually stated in two ways:

• As a percentage of the universe. “An estimated 64 percent of TV households in the market” means 64 of every 100 households in that market saw the PSA at least once.
• As a number of people or households. Multiply the percentage by the universe estimate. 64 percent of 640,000 households is about 409,600 households.

Say who is being counted. Household reach and reach among adults 25 to 54 are different numbers from the same schedule, and a report should name the one it shows.

Reach also belongs to a market and a period. “Estimated reach in the Denver DMA, March 1 to May 31” is a complete figure. “Estimated reach” on its own isn’t.

What is frequency?

Average frequency is the number of times the average reached person was exposed. If a schedule reached 400,000 households and delivered 1.2 million household impressions, the average household that saw it saw it three times.

An average hides the spread. Some households see a PSA once, and a few see it ten times. So planners also look at 3+ reach, the share of the universe exposed at least three times. Many planners treat three exposures as a working minimum for a message to register, though the right number depends on the message. A simple awareness message can land in one or two exposures. A message that asks people to change a habit usually needs more.

How do reach, frequency and GRP fit together?

Gross rating points (GRP) add up the ratings of all the airings in a schedule. They count exposures, not people. Reach and frequency split those exposures into how many people were reached and how often.

GRP                = Reach (%) × Average Frequency
Average Frequency  = GRP ÷ Reach (%)
Reach (people)     = Reach (%) ÷ 100 × Universe Estimate
Gross Impressions  = GRP ÷ 100 × Universe Estimate

A schedule of 300 GRP that reaches an estimated 75 percent of a market has an average frequency of 4.0. In a market of 640,000 households, that’s about 480,000 households reached and 1,920,000 gross household impressions. All four numbers describe the same schedule; they answer different questions. The GRP calculator works out GRP from impressions and a universe.

Why can’t reach be worked out by adding up airings?

Because the same people see more than one airing. Gross impressions count every exposure, so a household that saw a PSA five times counts five times. Reach counts it once. That’s why reach is always smaller than gross impressions, and why it grows more slowly than GRP.

The relationship is a curve. A schedule’s first hundred or so GRPs reach a large share of the market quickly, because most exposures land on people who haven’t seen the spot yet. After that, more and more exposures land on people who already have, so extra GRPs add a little new reach and a lot of frequency. Doubling a schedule’s GRPs doesn’t come close to doubling its reach.

Nobody records which households saw which airing, so broadcast reach can’t be counted directly. It’s estimated from the schedule and the audience data behind it. That’s why PSAReach labels it “estimated” wherever it appears.

What changes reach for the same number of GRPs?

Two schedules with the same GRPs can reach very different numbers of people. What matters is how the airings are spread.

• Stations. Airings spread across several stations reach more different people than the same GRPs on one station, where the same loyal viewers or listeners see the spot again and again.
• Dayparts. Morning news, daytime, prime and late night draw partly different audiences. A mix of dayparts builds reach; repeating one daypart builds frequency.
• Flight length. A schedule spread over eight weeks meets more new people than the same airings packed into one week, because audiences turn over.
• Rating size. Donated airings often run in lower-audience periods, so each one carries a small rating. Reach still builds. It takes more airings to get there.

These are the same levers a planner pulls on purpose. If the goal is reach, spread the schedule. If the goal is frequency among one audience, concentrate it.

How should reach be reported across markets and media?

A few rules keep reach figures sound once a report leaves the analyst’s desk.

• Report reach per market and per period, labeled estimated. A reach figure without its market, period and audience can’t be checked or compared.
• Add markets, not weeks. Each county belongs to one DMA, so the people in two markets are different people, and market reach adds up to a national figure. Weekly reach in one market doesn’t add up, because most of the same people appear in each week.
• Keep TV and radio apart. The overlap between the TV viewers and the radio listeners of one campaign can’t be measured, so each medium’s reach is reported on its own.
• Name the source. Broadcast reach and frequency are estimated. Reach and frequency for a paid digital campaign come from the advertising platform’s own reporting, so they’re measured, with the platform named as the source.

What are reach and frequency used for?

Reporting to a board or funder. “An estimated 580,000 households in two markets saw the PSA, about three times each on average” is a sentence a board member can repeat. An airing count isn’t.

Setting goals before a campaign. A goal such as “60 percent of adults 25 to 54 in our five priority markets, three or more times” gives the plan a target that airings alone can’t express.

Finding gaps. A market with high frequency and low reach is seeing the same people over and over; the usual fix is more stations or more dayparts. A market with low reach and low frequency simply didn’t get enough airings, and it’s the natural place for paid media to fill in.

Comparing donated and paid media. Reach and frequency put donated broadcast, paid broadcast and paid digital on the same terms, so you can see what each one added. The guide to combining donated and paid media goes further.

Planning the next campaign. Reach and frequency by market, over several campaigns, show which markets and stations carry your PSAs well and which need a different approach.

How does PSAReach report reach and frequency?

PSAReach campaign reports include, for TV and radio separately:

• estimated reach by market and nationally
• estimated average frequency
• estimated 3+ reach
• the GRP and gross impressions behind them, next to the measured airings

Each figure carries the “estimated” label and sits below the measured airing counts. The same figures appear in AdModus dashboards and exports. If you’d like to understand how we produce them for your campaign, get in touch and we’ll walk you through it.

A worked example

Suppose a PSA runs for six weeks in two markets. The numbers are invented for illustration.

Market A:  Average Frequency  = 180 ÷ 64 = 2.8
           Reach (households) = 64 ÷ 100 × 640,000 = 409,600

Market B:  Average Frequency  = 90 ÷ 42 = 2.1
           Reach (households) = 42 ÷ 100 × 410,000 = 172,200

The two markets are separate populations, so their reach adds up: an estimated 581,800 households across both. Across both markets the schedule delivered about 1,521,000 gross household impressions, so the average frequency is about 2.6.

The report line reads: “The PSA reached an estimated 582,000 TV households in two markets, about 2.6 times each on average.” The measured airing count sits right above it.

For the next flight, Market B is the one to look at. Three stations and 42 percent reach suggest adding stations or a paid flight there, while Market A mostly needs the same again.

A checklist for reach and frequency figures

1. Reach is stated per market and per period, and the audience is named (households or a named age group).
2. Broadcast reach and frequency carry the word “estimated”.
3. Market reach is added into a national figure; weekly reach in one market isn’t.
4. TV and radio reach are reported separately.
5. GRP, reach and frequency agree: GRP equals reach times average frequency.
6. Figures are rounded to the precision the audience data can support.

For the definitions behind this guide, see reach and frequency in the glossary. The guide to measured and estimated impressions explains how the impressions underneath these figures are built and labeled.