# How PSA tracking works for TV and radio

> A plain-language guide to how PSA airings on TV and radio are detected, verified, normalized and turned into campaign reports nonprofits can trust.

Source: https://psareach.com/insights/how-psa-tracking-works-tv-radio/

PSA tracking Guide · 8 min read

A plain-language guide to how PSA airings on TV and radio are detected, verified, normalized and turned into campaign reports nonprofits can trust.

 The PSAReach team Published September 23, 2026 · Reviewed September 29, 2026

When a nonprofit buys advertising, the invoice tells it what ran. Donated airtime has no invoice. A station that airs a public service announcement decides when, how often and for how long, and it rarely sends a schedule. PSA tracking closes that gap. It turns a campaign that “went out to stations” into a verified record of what actually aired, where and when.

This guide takes each step of that process in plain terms. It’s written for communications and program directors who need to know what a tracking report can and can’t tell them, with enough detail for the analyst who has to stand behind the numbers.

## Why is tracking different for donated media?

Paid advertising is planned in advance. The buyer picks the station, the program and the number of spots, and the station confirms delivery with an affidavit. Donated media works the other way around. Stations get PSA creative from many organizations and fit it into unsold inventory as they see fit.

Some will run a spot dozens of times in overnight hours. Others will put it in a morning newscast once a week. A few won’t air it at all.

So a PSA campaign has no plan to compare against. The only way to know what happened is to observe the broadcast itself. Tracking is that observation layer, and the rest of a campaign report, from impressions to media value, is built on top of it.

## Step 1: give each creative an identifier

Tracking starts before the first airing. Each version of a PSA needs a unique code so an airing can be matched back to the right campaign, message and length. In U.S. broadcast, that code is an Ad-ID. A campaign with a :30 and a :60 in English and Spanish has four creatives, so it needs four codes.

The code matters for another reason. The same message often exists in several cuts, and a tracking report that lumps a :15 and a :60 together will misstate value, because value depends on length. With a code on each cut, the tracking system keeps them apart from the start.

### What goes wrong without identifiers?

Without a unique code, a monitoring service has to match airings by comparing the audio or video with a reference file. That still works, but revisions and re-edits get hard to tell apart. A campaign that quietly swapped its end card mid-flight can end up with two partial records instead of one complete one. Assign codes early and put them on the slate of each delivered file.

## Step 2: detect the airings

Broadcast monitoring services detect airings by recording station signals across U.S. markets. The most common methods are audio and video fingerprinting, where the monitoring system compares what aired with a reference copy of the creative, and encoded watermarks that the creative carries in its audio. Both produce an occurrence record: the station, the date, the local time and the creative that matched.

Coverage is the first thing to ask any tracking service about. Monitoring covers a large share of U.S. television stations and a large share of commercial radio, but not every signal in every market. A station outside the monitored set can air a PSA every day and the report will show nothing. A good tracking provider tells you which stations it monitors and flags markets where coverage is partial.

TV and radio differ in practice. TV occurrence data tends to be more complete, because there are fewer stations and monitoring is denser. Radio has far more stations, formats change, and low-power and noncommercial signals may not be monitored. Radio reports carry more coverage caveats as a result, and analysts should spell them out.

## Step 3: normalize the record

A raw occurrence file isn’t a report yet. Each detection has to be tied to a canonical station record, placed in a market, converted to a local daypart and checked for duplicates. It’s the least visible step, and it’s where things most often go wrong.

-   **Station identity.** Call signs change, stations switch networks and owners, and the same call sign can turn up with different suffixes in different files. Normalization maps each detection to one station record with its network or format, owner, city and market.
-   **Market.** TV airings are placed in a DMA, the standard U.S. television market designation. Radio airings go in a radio metro. Markets don’t overlap, so a national campaign’s coverage can be added up market by market without double counting.
-   **Daypart.** The airing time is converted to the station’s local time zone and assigned to a standard daypart such as early morning, daytime, prime or overnight. That’s what lets you judge the campaign’s placement quality later.
-   **Duplicates.** Two monitoring sources can report the same airing, and simulcasts or repeated feeds can make one placement look like several. Write the merge rules down and apply them consistently.

## Step 4: enrich the record

Once the airing record is clean, it can be enriched with audience and value estimates. This is where the report moves from measured to estimated, and the labels have to change with it.

Audience enrichment attaches an estimated number of households or people to each airing. For TV, that usually means a rating for the program or daypart multiplied by the market’s universe estimate. For radio, it means an average quarter-hour audience for the daypart. Neither figure counts who saw that particular airing. Both are estimates drawn from audience research, and their source period should appear in the report.

Value enrichment attaches an estimated commercial-equivalent value to each airing. It answers “what would this placement have cost a commercial advertiser?” using rate evidence for the same station, daypart and spot length. That’s an estimate too, so a good report labels it estimated and names the rate basis.

## Step 5: report what was measured and what was estimated

A responsible PSA tracking report keeps the two kinds of numbers apart. Here’s the usual split.

| Figure | Basis | Where it comes from |
| --- | --- | --- |
| Airings, stations, markets, dayparts | Measured | Measured occurrence data, normalized |
| Creative and spot-length breakdown | Measured | Identifier match on each airing |
| Impressions | Estimated | Ratings or AQH times a universe estimate |
| Commercial-equivalent value | Estimated | Rate evidence by station, daypart and length |
| GRP and TRP | Estimated | Derived from estimated impressions |
| Reach and frequency | Estimated | Modeled from the schedule within each market |

PSAReach reports follow this split on every page. We state measured figures plainly. Estimated figures carry the word “estimated”, the method used and the period of the underlying source data. The [methodology page](https://psareach.com/methodology/) explains each calculation.

## A worked example

Suppose a campaign sends a :30 and a :60 to stations nationally and runs for eight weeks. At the end of the flight, the tracking file shows 1,240 verified airings on 86 stations across 41 DMAs. Of those, 890 are the :30 and 350 are the :60. About 38 percent fall in overnight hours, 27 percent in daytime and 22 percent in early morning news, with the rest in evening and prime.

Those numbers are measured. They come straight from the normalized occurrence data, and the program director can quote them without qualification.

Then comes enrichment. Each airing is matched to an audience estimate for its station and daypart, and the total comes to, say, 9.6 million estimated gross impressions. Each airing is also matched to a commercial-equivalent rate, which adds up to an estimated donated media value of $410,000. The :60 airings are worth more per airing than the :30s, but they don’t deliver more impressions, because a longer spot in the same break reaches the same people.

Divide estimated gross impressions by the relevant universe in each market and you get estimated GRP per market, which can be added across the eight weeks. Reach works differently. A modeled reach of 2.1 million people in one market can’t be added to next week’s reach in the same market, because many of the same people would be counted twice. It can be added across markets, since a household in one DMA isn’t in another.

## What can’t tracking tell you?

Even a careful tracking report has limits. It’s better to state them up front than to have a board member find them.

-   It can’t report airings on stations outside the monitored set. A station’s absence from the report doesn’t prove it didn’t air the spot.
-   It can’t count the viewers of one specific airing. Audience figures are estimates for the program or daypart, not observations of that break.
-   It can’t produce a measured unduplicated reach. Reach across a schedule is modeled, and cross-media reach across TV, radio and digital shouldn’t be combined into one figure.
-   It can’t tell you whether the airing changed behavior. Linking airings to search or donation activity is a separate analysis with its own caveats.

## What should you ask your provider?

Use these questions when you evaluate any PSA tracking service, including [PSAReach](https://psareach.com/psa-tracking-analytics/).

1.  Which TV and radio stations do you monitor, and which markets have partial coverage?
2.  How do you detect airings, and how do you handle a creative that was revised mid-flight?
3.  How do you remove duplicate detections from simulcasts and multiple feeds?
4.  Can I get a station-level export with local airing time and daypart for each airing?
5.  Which figures in the report are measured and which are estimated, and is that labeled on every page?
6.  What audience source and period sit behind the impression estimates?
7.  Is media value labeled estimated, and what rate evidence is it based on?
8.  Do you treat :15, :30 and :60 airings separately for value, and do you avoid scaling audience by length?
9.  Do you sum reach across weeks, or report it per period and per market?
10.  Can the report be refreshed on a schedule, and does each refresh keep its source dates?

Whatever a nonprofit claims about its PSA campaign rests on the tracking. Get the airing record right and label the estimates clearly, and the rest of the report will hold up. For definitions of the terms used here, see the [glossary](https://psareach.com/glossary/#term-airing). For how PSAReach turns airings into audience and value intelligence, see [Media Intelligence](https://psareach.com/media-intelligence/).

Terms in this guide

-   [PSA (public service announcement)](https://psareach.com/glossary/#term-psa)
-   [Airing](https://psareach.com/glossary/#term-airing)
-   [Station](https://psareach.com/glossary/#term-station)
-   [DMA (Designated Market Area)](https://psareach.com/glossary/#term-dma)
-   [Daypart](https://psareach.com/glossary/#term-daypart)
-   [Ad-ID](https://psareach.com/glossary/#term-ad-id)
-   [Spot length](https://psareach.com/glossary/#term-spot-length)
-   [Impressions](https://psareach.com/glossary/#term-impressions)
-   [Estimated vs measured](https://psareach.com/glossary/#term-estimated-vs-measured)

[See all guides](https://psareach.com/insights/)

## Get new guides by email

One short email a month with new guides on donated media, PSA tracking and campaign reporting. Reading the guides never requires an email address.

Check your email to confirm. Nothing is sent to you until you do.

Related services

-   [PSA Tracking & Valuation Verified TV and radio airings by market, station and daypart, with the airtime valued and labeled estimated. Explore](https://psareach.com/psa-tracking-analytics/)
-   [Dashboards & Reporting Executive dashboards and automated reporting for boards, funders and program teams. Explore](https://psareach.com/dashboards-reporting/)
-   [Media Intelligence Station, market and audience insight, backed by a transparent valuation methodology. Explore](https://psareach.com/media-intelligence/)

## Want these numbers for your own campaign?

The guides show the method. Tell us about your public service announcements and we'll apply it to your campaign.

[Talk to PSAReach](https://psareach.com/contact/) [Explore all services](https://psareach.com/#services)