Key message pull-through: Definition, origins, and calculation

In this article

Key message pull-through is the percentage of analyzed media coverage that contains a predefined key message with its required meaning intact. If 44 out of 80 relevant articles carry a message, its pull-through is 55%.

The metric answers one precise question: did the intended message make it into the coverage? It does not prove that the target audience saw the article, understood the message, or believed it.

What is key message pull-through?

Before a campaign, PR teams decide which small set of claims or ideas they want the audience to take away. These might describe the organization's position on an issue, a product benefit, or an important fact. After publication, the team reviews each relevant item of coverage to see whether those messages appear accurately or as faithful paraphrases.

Practitioners also use terms such as message penetration, key message delivery, and message inclusion. These labels are not defined consistently across every organization. A report should therefore state exactly how its number was calculated.

A practical definition is:

Key message pull-through is the percentage of relevant, analyzed media items that contain a specified key message.

Where does the metric come from?

Key message pull-through has no single inventor or universally accepted launch date. It comes from media content analysis, where communication practitioners code characteristics of coverage such as topic, tone, spokesperson presence, and the inclusion of intended messages.

A 2012 proposal from the Institute for Public Relations uses message penetration for the percentage of items containing one or more key messages. A 2017 study uses message delivery and also examines omitted, negative, or erroneous messages. Message pull-through is a common practitioner label today, but it is not one universally standardized formula.

The AMEC Integrated Evaluation Framework organizes communication measurement into connected stages. The presence of a message in a media item describes an output and the content quality of that output. Whether people recall or understand the message is an outtake. A change in attitude or behavior is an outcome.

That distinction matters because "the message got through" is often used for two different claims:

Pull-through answers only the first question. That is still valuable, as long as the report does not claim more.

The formula

Calculate pull-through separately for each key message:

Key message pull-through = Relevant items containing the full, accurate message ÷ relevant items to which that message applies × 100

The numerator counts media items in which the selected message appears at least once with its required meaning. The denominator includes only items to which that message applies. A message about one product, for example, does not belong in the denominator for coverage about another product from the same company.

For a campaign overview, you can also report the percentage of all relevant items that contain at least one complete key message. Label it any-message pull-through or message penetration so readers do not confuse it with the result for one specific message.

If one article repeats the same message five times, it still counts once in an article-level analysis. Counting individual statements or paragraphs measures message frequency instead. That is a different metric and needs a different denominator.

A worked example

A fictional software provider analyzes 80 deduplicated media items about a product launch. The team defined three key messages before the campaign began.

The team can immediately see which idea appeared often and which one received little coverage. The comparison does not explain why. That requires reading the coverage and checking plausible reasons. Was the third message too abstract? Was it buried at the bottom of the press release? Did spokespeople leave it out of interviews? Was it simply uninteresting to journalists?

An additional figure such as "65% of coverage contained at least one key message" may be useful. It should not replace the result for each individual message. Otherwise, one strong message hides the weak ones.

How to make the measurement reproducible

The formula is simple. Most of the work lies in the rules set before coding begins.

1. Define the messages first

Write each key message as a clear claim. A single theme such as "innovation" is not enough. Two people should be able to decide whether an article contains the claim.

2. Define the denominator

Set the period, markets, languages, media types, and topic criteria. Decide how to handle wire stories, syndicated copies, updated articles, and duplicates. Pull-through changes when the denominator changes.

3. Create a codebook

For each key message, the codebook should contain:

• the full claim,
• acceptable paraphrases,
• examples that count,
• similar claims that do not count,
• rules for partial, erroneous, and contradictory versions,
• the topics and item types to which the message applies.

A message does not need to appear word for word. A faithful paraphrase can count when it preserves the meaning. A related topic without the claim should not count. Partial or incorrect versions should remain visible, but they should not enter the same rate as complete messages.

4. Record the unit of analysis

The media item is usually the most useful unit for PR reporting. Code each article with a yes or no for every message. If one article can contain several messages, use a multi-select field rather than forcing one choice.

5. Review ambiguous cases

For important studies, have two people code a sample independently. Disagreements reveal where the codebook is vague. Automated or AI-assisted classification can speed up the process, but it should follow the same rules and be checked against a human-reviewed sample.

What is a good pull-through rate?

There is no credible universal benchmark. Whether 55% is good depends on the objective, message, media mix, and baseline. A short message that matters to journalists is more likely to appear than a complicated corporate position.

Useful comparisons include:

• the same message before and after a campaign,
• several messages within one campaign,
• target publications versus all coverage,
• interviews and background briefings versus press-release pickups,
• markets or languages analyzed with the same coding rules.

Report the absolute figures alongside the percentage. "44 of 80 items" says more than 55% without a sample size.

What the metric cannot show

A high pull-through rate is not automatic proof of communication success.

• No verified reach: Message presence does not show how many people saw the article.
• No proof of understanding: Only a survey or a comparable test can show whether the audience understood the message.
• No agreement: A message may appear in a critical or hostile context. Measure tone and framing separately.
• No behavioral impact: Leads, applications, donations, or policy support need their own outcome measures.
• No causal proof: A higher result does not by itself identify which PR activity caused the change.

The Barcelona Principles 4.0 call for qualitative and quantitative analysis and measurement beyond outputs. Pull-through is one useful part of that work, not a substitute for outtakes, outcomes, or impact.

How to measure key message pull-through in aclipp

In aclipp, create a multi-select metric called "Key messages." Each predefined message becomes an option. When reviewing a clipping, select every message that appears according to the codebook.

The report then needs two figures for each message:

1. the number of relevant clippings containing that message,
2. the total number of relevant clippings under the same filter.

Keep the period, campaign, sources, and duplicate rules consistent. You can then compare pull-through over time or between defined groups.

Four details belong in every report

When a report states a pull-through rate of 55%, the reader should be able to find four details: the exact message, the denominator, the unit of analysis, and the coding rule. Without them, the number cannot be reproduced.

For the broader measurement context, read our guide to PR KPIs in 2026. Sentiment analysis separates message presence from tone, while AI visibility measurement applies message accuracy to generated answers.