Earned Media Statistics for 2026 - Clear Spark Digital

Earned media is any coverage of you or your work that appears in a place you don’t own and didn’t pay for. A quote in a trade article, a podcast interview, a feature about you or your business, or a reference to data that you published. What that coverage is worth has changed in the last two years.

The short version is that referral clicks from a placement are declining quickly, and AI citations from that same placement are rising.

The key earned media statistics

Earned media accounts for about 84% of the links generative AI models cite, while owned and paid content together make up roughly 16%.
Journalism alone accounts for 27% of AI citations.
57% of AI-cited journalism articles with a known publish date were published in the previous 12 months.
Paid and advertorial content accounts for just 0.3% of AI citations.
Press releases make up 1.1% of all AI citations.
Google organic search traffic to more than 2,500 news sites fell 33% globally and 38% in the US between November 2024 and November 2025.
When a Google AI summary appears, users click a traditional search result 8% of the time, compared with 15% when no summary appears. They click a link inside the summary itself just 1% of the time.
66% of journalists rely on PR-provided material for story ideas, making PR the leading source of story leads.
86% of journalists will reject a pitch that isn’t relevant to their beat or audience, and most receive 50 or more pitches a week.
Trust in news has fallen to 37% globally, the lowest level since the Reuters Institute began measuring in 2015.
66.2% of digital PR practitioners now track AI citations as a KPI, a metric that did not appear in the previous year’s survey.

What AI Cites When It Answers a Question About Your Industry

When someone asks ChatGPT about your category, competitor, or product, the answer comes from a source somewhere. Muck Rack’s May 2026 edition of What Is AI Reading? analyzed more than 25 million links cited across ChatGPT, Claude, and Gemini responses spanning 17 industries. The finding: 84% of those citations came from earned media. Owned and paid content combined for the other 16%. Journalism alone accounted for about 27%.

Those numbers have held steady for a while. Across three editions of the study going back to July 2025, earned media has ranged from 82% to 89% of citations, and journalism has stayed between 25% and 27%. The consistency suggests the citation behavior isn’t a quirk of one model update. It’s how these systems have decided to source information.

Press releases, meanwhile, account for just 1.1% of citations. Wire distribution generally isn’t an effective AI-visibility strategy (although, of course, press releases serve other purposes).

A different research group reached a similar conclusion using a completely different method. Fullintel, working with the University of Connecticut and presenting at the International Public Relations Research Conference in February 2026, ran 400 prompts across 10 personas covering weight-loss medication queries. Their content analysis found 89%+ of cited links came from earned media and roughly 95% from unpaid sources. Different researchers, different topic, different methodology, but very similar findings.

A few nuances from the Muck Rack data worth noting:

Query type matters. Industry trend questions produce journalism citations in 46% of responses, more than double the rate for how-to and comparison queries.
Freshness matters. 57% of AI-cited journalism with a known publish date was published in the previous 12 months. This shows that showing up in AI citations is an ongoing battle.
The long tail of outlets is where most citations sit. AI models cite journalism across more than 20,000 distinct outlets. Axios is the only publication to break into ChatGPT’s top three cited domains in 13 of 17 industries.
Model behavior isn’t uniform. ChatGPT cites sources in 96% of responses, averaging five per response. Gemini cites in 82% of responses, averaging eight. Claude cites in 55%, but averages 13 when it does.

A March 2026 Stacker GEO study found that distributing the same story across third-party news outlets produced a median 239% lift in AI citations versus keeping it on the brand’s own site alone. An earlier Stacker and Scrunch test saw citation rates rise from about 8% to 34% when brand-owned content moved onto third-party outlets, across 944 prompt-and-platform combinations.

AI citation behavior can shift when models are retrained. Muck Rack says as much in its own methodology. These figures are a snapshot of mid-2026.

The Referral Traffic Your Coverage Used to Send is Disappearing

If you measure a placement by the visitors it drives to your site, coverage probably looks like it stopped working. That’s because the mechanism that used to send those visitors is falling apart, independent of anything you’re doing.

Reuters Institute research published in early 2026 found that Google organic search traffic to more than 2,500 news publishers fell 33% globally between November 2024 and November 2025. In the US, the drop was 38%. Traffic from Google Discover fell 21% year over year. Publishers surveyed by the Institute expect search traffic to fall a further 43% over the next three years.

That means the outlet that ran your feature is sending you significantly fewer readers than it did a year and a half ago.

On the search side, a Pew Research Center study of 900 US adults tracking 68,879 Google searches in March 2025 found that when an AI-generated summary appeared on the results page, users clicked a traditional result 8% of the time. When no summary appeared, they clicked 15% of the time. Clicks on links inside the AI summary itself happened in about 1% of visits. Users were also more likely to end their browsing session entirely after a page with a summary (26%) than without one (16%). Roughly one in five searches in that dataset produced a summary. Pew is careful to note that this shows association, not causation.

The old case for coverage was that it sent qualified traffic to your website. That case is weakening.

The new case is that the same coverage now feeds the systems answering questions for people who never see a search result at all. If you’re measuring only what your analytics platform can see, you’re measuring the half of the value that’s shrinking and missing the half that’s growing.

What People Trust in 2026

Trust in news itself is at a record low. The 2026 Reuters Institute Digital News Report, which surveyed nearly 100,000 news consumers across 48 markets, put global trust in news at 37%. That’s the lowest reading since the Institute began measuring in 2015. Trust declined in half the markets surveyed, with drops of five percentage points or more in 19 of the 48. Hungary posted the lowest reading at 17%, followed by Greece at 18%.

Where people encounter news matters as much as whether they trust it. For the first time, social media and video platforms passed news organizations’ own websites as the most widely used source of news globally, at 54% versus 51%.

Trust varies sharply by channel. On a global level, 32% trust news in search engines, 22% trust news on social media, and 20% trust news from AI chatbots. Among people who actually use AI chatbots for news, though, trust rises to 44%, according to the Reuters Institute’s chapter on AI chatbots and news. Weekly use of AI chatbots for news went from 7% to 10% globally over the past year, and reaches 17% among 18 to 24-year-olds. Only 1% call chatbots their primary news source.

All of which is context for the more practical question: when you land coverage, whose audience are you actually reaching?

Pew Research Center surveyed 9,482 US adults about 30 major US news sources, with a margin of error of plus or minus 1.4 percentage points. A few findings that are useful when you’re picking outlets:

Of those 30 sources, the average Forbes reader sits closest to the party identification and ideology of the average US adult. If a politically neutral audience matters for what you’re placing, that’s a data-backed pick.
Awareness varies enormously. 94% of US adults have heard of CNN. 78% have heard of Forbes. 66% have heard of the Associated Press. Just 30% have heard of Axios.
Trust in the largest cable outlets is intensely partisan. 58% of Democrats say they trust CNN while 58% of Republicans distrust it. 56% of Republicans trust Fox News while 64% of Democrats distrust it.

Compare the Axios awareness number (30%) with the earlier finding that Axios is the most-cited news outlet in ChatGPT responses across 13 of 17 industries. The publication most Americans have never heard of is the one the machines cite the most.

What Journalists Want

Two large 2026 surveys of working journalists landed within two months of each other. Both are vendor-published, both have transparent methodology, and together they cover more than 2,700 reporters.

Cision’s 2026 State of the Media Report surveyed 1,899 journalists across 19 markets in January and February 2026. The headline is that 66% of journalists rely on PR-provided content, including releases, pitches, and media kits for story ideas. That makes PR the single leading source of story leads for the newsroom.

The rest of the report describes what makes a pitch land or die:

Most journalists receive 50 or more pitches a week, according to Cision’s summary, and most aren’t relevant to the reporter’s beat.
86% say they will reject a pitch that lacks relevance to their audience or beat. 79% say they’re more likely to engage with a relevant pitch.
What journalists say they value most in a pitch: original research, expert access, and embargoed information.
53% oppose AI-generated pitches, citing accuracy and personalization.
50% name accuracy, fact-checking, and combating misinformation as their biggest challenge, up from 40% the year before. 49% cite shrinking budgets and heavier workloads, up from 29%. 43% rank the rise of AI among their biggest challenges, up from 30%.
62% use LinkedIn professionally, and 33% rank it as their single most valuable platform.

Muck Rack’s 2026 State of Journalism report surveyed 1,044 journalists between January 30 and March 2, 2026, with 897 responses included after data quality checks. A few figures worth knowing:

82% of journalists report using at least one AI tool, up from 77% the previous year. ChatGPT usage climbed to 47%, Gemini to 22%, Claude to 12%.
26% cite unchecked AI as a top industry concern, up from 18%. Disinformation and lack of funding tie as the top concerns at 32% each.
LinkedIn is the most trusted social platform among journalists (58%). Distrust in TikTok climbed to 61%.
Only 3% of journalists have their stories assigned by an editor. 40% choose the stories they want to write.

How Much Earned Media Matters In Your Industry

The headline figures on AI citations hide a wide range across categories. The Muck Rack May 2026 report breaks its data down by 17 industries. Media and entertainment queries return journalism citations at 36.3%, the highest of any sector. Finance and insurance sits second at 32.1%. Industrial and manufacturing queries sit lowest at 18.6%. The study average for journalism citations is about 27%.

Manufacturing is the interesting counterexample. In that category, 36.6% of AI citations go to corporate blogs and owned content, which is well above the 24.0% study average. If you’re in a heavy-industry vertical, your own website content may actually be doing more of the citation work than a services or consumer category would suggest.

A high journalism citation rate means AI systems are answering questions in your category by reaching for editorial coverage. That’s the case for an earned media investment. A high owned-content rate means your own site can still compete on its own. In practice, most B2B and professional-services categories fall closer to the media-and-entertainment end than the manufacturing end.

What Practitioners Are Spending and Measuring

BuzzStream’s State of Digital PR 2026 report surveyed roughly 150 digital PR practitioners. Its sample skews UK-based (57%), so read it as directional for the US market. The findings are still noteworthy:

68.2% of practitioners say digital PR is more effective than it was a year earlier, roughly a 20-point jump from the prior survey.
Monthly digital PR budgets above $20,000 more than doubled year over year, from 4% of respondents to 8.8%. Sub-$5,000 budgets fell from 34.1% to 25.7%.
66.2% now track AI citations as a KPI. That metric didn’t appear in the previous year’s survey at all.
39.2% cannot name their own average cost per link, down from 51.4% a year earlier.

The budget-mix and KPI shifts are worth considering together. In a single year, the median practitioner has moved from being unsure what a link costs and not measuring AI citations at all, to tracking AI citations as a headline metric and spending materially more per month to earn them. The measurement question is being answered in real time by the field itself.

The Short Version

Earned coverage sends fewer clicks than it did two years ago. That decline is real, and it isn’t reversing.

Earned coverage also carries a form of value it didn’t carry two years ago: it’s what generative AI systems rely on when they answer questions about your category, at a rate of 84%. Those AI answers are being consumed by an audience your analytics dashboard probably cannot see.

If your PR program is measured on referral traffic alone, it will look worse each quarter. If it’s measured on which coverage gets cited when a customer asks an AI about you, the picture is almost the opposite. This shift has changed how the industry measures impact. Two-thirds of practitioners are now tracking AI citations. A year ago, essentially nobody was.