AI Search Now Rewards Data and Earned Media Over Marketing Copy

AI search engines are increasingly citing statistics, independent studies and journalism rather than brand-written marketing copy. A Princeton study found that adding statistics to content lifted visibility in AI search answers by 30 to 40% in benchmark tests, while separate industry research shows journalism and earned media now drive a large share of what AI platforms choose to mention. For founders, the message is blunt: AI systems trust third-party proof far more than they trust what your own website says about you.

What changed?

A report published this month by PR firm Wakefield Research set out new evidence on how large language models decide which brands to mention. The research draws on a Princeton paper which found that adding statistics was among the most effective generative engine optimisation (GEO) techniques, with top-performing methods improving content visibility by 30 to 40% in benchmark tests, according to O'Dwyer's PR News.

The underlying argument is that ChatGPT, Gemini, Claude, Perplexity and other AI platforms are not simply ranking websites the way traditional search engines do. They are pulling information from across the web and surfacing what they judge to be the most authoritative, credible and well-supported sources, as reported by O'Dwyer's PR News.

That shift is why PR and communications specialists are now being pulled into what used to be a purely technical SEO conversation. As large language models increase the value they place on trusted third-party sources, data, research and earned media are becoming powerful drivers of brand visibility, putting PR and comms teams in a stronger position to shape how brands are discovered.

Separately, a new citation study from agency 5W has mapped exactly which sources AI platforms lean on. It found that journalism accounts for 27% of all AI citations, rising to 49% on time-sensitive queries, according to 5W's AI Platform Citation Source Index. The same study warns that citation share can be volatile within weeks rather than years: ChatGPT's citation share of Reddit fell from roughly 60% to 10% in six weeks in late 2025, after a single Google parameter change.

Why does this matter for your business?

If your content strategy still leans on adjectives, this is the clearest signal yet that it needs to change. AI models are not impressed by claims of being "the leading" or "the best". They are built to look for verifiable, quotable proof: numbers, named studies, named experts and independent coverage.

The Wakefield report frames this as an opportunity rather than a threat. It states that this creates a significant opening for teams to lean into proprietary opinion research, producing credible, actionable insights that amplify thought leadership and fuel earned media coverage. In practice, that means original data, surveys and research reports are becoming genuine GEO assets, not just PR nice-to-haves.

There is a commercial angle too. A separate study from Semrush, which analysed brand visibility across AI platforms using 126 million AI search prompts, found a sharp performance gap based on how integrated a company's SEO and AI strategy is. Among organisations that fully integrate SEO and AI visibility into a unified workflow, 81% reported increased traffic or leads from AI platforms. Among organisations managing the two areas separately, only 36% reported the same result, according to Semrush. Companies treating AI visibility as someone else's job, whether that is PR, SEO or comms, are leaving results on the table.

The volatility point from the 5W study also matters commercially. If citation share can swing from 60% to 10% inside six weeks because of a single platform change, brands cannot treat a single AI visibility win as permanent. It needs ongoing monitoring, not a one-off audit.

What does the data show about AI citations?

The 5W index breaks citation behaviour down by platform, and the differences are stark enough to change how you prioritise effort. ChatGPT concentrates on Wikipedia, Reddit, Forbes and Business Insider, while Perplexity rewards primary sources, NIH and PubMed, and named B2B authority, according to 5W. A brand optimising purely for one platform's habits could be invisible on another.

Video is also becoming a bigger factor than many marketers realise. YouTube holds a 200x citation advantage over every other video source and dominates Google AI Overviews, according to the same research. If your brand has no video presence, you are conceding an entire citation category to competitors that do.

Put together with the Princeton findings on statistics, a pattern emerges. AI systems increasingly trust sources that combine three things: independent verification, specific numbers, and consistent presence across the platforms people actually use to ask questions.

What should you do now?

Start by auditing what proof actually exists on your key pages. If a page makes a claim about market position, customer results or product performance without a number, a source or a named study behind it, that page is a weak candidate for AI citation.

Next, look at where your brand is, and is not, picked up by earned media. Given that data, research and earned media are becoming powerful drivers of brand visibility, a single well-placed original research report, distributed properly, can do more for AI visibility than months of routine blog output.

Finally, treat AI visibility as a live, cross-platform measurement problem rather than a set-and-forget project. Given how quickly citation patterns can shift, as shown by the Reddit example in the 5W study, checking your own AI visibility on a regular basis, for instance with Sited's free audit, is a sensible habit rather than a one-time task.

Frequently asked questions

Does adding statistics to my website really improve AI search visibility?

According to research cited by Wakefield, adding statistics was among the most effective GEO techniques, with top-performing methods improving content visibility by 30 to 40% in benchmark tests. It is not a guarantee, but it is one of the more measurable levers available.

Why do ChatGPT and Perplexity cite different types of sources?

The platforms are built differently and prioritise different signals. ChatGPT concentrates on Wikipedia, Reddit, Forbes and Business Insider, while Perplexity rewards primary sources, NIH and PubMed, and named B2B authority, meaning a strategy tuned for one platform will not automatically work on another.

Is earned media more important than owned content for AI visibility now?

The evidence suggests earned coverage carries particular weight. Journalism accounts for 27% of all AI citations, rising to 49% on time-sensitive queries, so a strong PR and research programme can meaningfully boost how often AI systems mention your brand.

How stable is AI citation share once you achieve it?

Not very. Citation share is volatile within weeks rather than years; ChatGPT's Reddit citation share fell from roughly 60% to 10% in six weeks in late 2025 after a single Google parameter change. Ongoing monitoring is essential rather than a one-off check.

Does integrating SEO and AI visibility work actually produce better results?

Yes, according to Semrush's latest research. Among organisations that fully integrate SEO and AI visibility into a unified workflow, 81% reported increased traffic or leads from AI platforms, compared with only 36% among those managing the two separately.