Discussion starter: a digital PR question, posted with an answer from Niraj Raut to open the thread. If you have dealt with this on a site, reply with what you saw, especially where it differs.
- Digital PR
Journalists increasingly use AI tools to find sources and data. Has that changed which of your experts or studies get picked up?
Disclosure: Niraj Raut, who posted this answer, runs the SEO consultancy linked at the end of it.
1 reply
-
adminAdmin
It is changing what gets found, but not in a way anyone can yet measure cleanly. Journalists who use AI assistants or AI search to find a statistic or an expert are, in effect, adding another retrieval layer between your work and their story. That layer favours material that is easy to find, easy to understand out of context and easy to attribute: a clearly stated finding with a date, a sample and an owner, or an expert whose specialism is stated plainly in several places. Studies buried in PDFs and experts who exist mainly in a sales deck are the ones that lose out. What nobody has published is a clean before-and-after of pickup rates, so treat any claimed percentage with suspicion, including your own until you have tracked it properly.
How the discovery path has shifted
The old path was a search, a skim of the first page of results, and perhaps a request posted to a sourcing service. The newer path often starts with a question to an assistant: “what share of UK renters have been asked for a guarantor?” or “who are credible experts on warehouse automation?” The assistant retrieves pages, summarises and sometimes names sources. The journalist then checks the original, if they check at all.
Three consequences follow for PR and SEO teams:
- The page that holds the finding matters more than the press release. The assistant is more likely to retrieve a well-structured findings page on your domain, or coverage of it on an established publication, than a release on a wire service.
- Attribution can drift. If a larger site re-reported your figure, the assistant may name them instead of you. Journalists then credit the second-hand source.
- Expert discoverability depends on a public footprint. Assistants can only suggest people who appear in retrievable sources: bylines, quotes in previous coverage, conference pages, clear bio pages. An excellent internal expert with no footprint is invisible to that path.
What makes a study easier for both journalists and assistants to use
Element Why it helps Headline findings as short, complete sentences near the top of an HTML page Each finding can be lifted and understood on its own, with the number, population and date in the same sentence Method section: sample size, dates, who ran it, how respondents were recruited Journalists need this to use the figure; its absence is a common reason a stat is dropped Named owner of the research and a contact route Gives the journalist a person to quote and the assistant an entity to attribute Data tables in HTML, with a downloadable file as a secondary option Figures in images or PDFs are harder to retrieve and quote accurately A stable URL that is updated when the study is repeated, with past editions kept Links and citations accumulate in one place instead of splitting across yearly URLs A suggested citation line Reduces misattribution when the figure travels None of these guarantee a citation in an AI answer. They remove reasons not to use your work, which is the part you control.
Making experts findable
For experts, the practical work is building a consistent public record of what each person is qualified to talk about.
- A bio page per expert on your site, stating their specialisms in plain words, with credentials, previous media appearances and topics they comment on.
- Consistent descriptions across the bio page, LinkedIn, conference listings and author pages, so an assistant sees the same person described the same way.
- Bylined articles on your site and in trade press on the specific topics you want them associated with. Breadth dilutes; three topics are more useful than fifteen.
- Quotes in coverage that name their organisation and role correctly. Ask for corrections when a publication gets them wrong.
Watch for this:
Journalists have become wary of expert responses that read as AI-generated, a problem that grew with sourcing platforms. Short, generic comments now hurt an expert’s reputation with the journalists who notice. Specific, experience-based answers, delivered quickly and verifiably by the named person, stand out more than they used to.
How to find out whether it is changing your pickups
The trouble with this question is that very few teams recorded how journalists found them before assistants arrived, so there is no baseline. Start one now.
- Ask. When a journalist gets in touch, ask how they found the study or the expert. Log the answer in the coverage tracker. After two or three quarters you will have a real distribution.
- Watch referrals to study pages through GA4’s AI Assistant channel. A rise in assistant referrals to research pages is a sign that people, some of them journalists, are arriving that way.
- Test the prompts journalists would use. Ask several assistants for the statistic your study answers and for experts on your topic. Record whether you appear, who appears instead and whether attribution is correct. Repeat quarterly with the same prompts.
- Track unlinked and misattributed mentions of your figures. A growing share of coverage that uses your number but credits someone else is a signal that the retrieval layer is picking up a re-report rather than your original.
Hypothetical example:
A lender publishes an annual survey as a PDF with a press release. A comparison site writes it up with better formatting, and when journalists ask an assistant for the statistic, the comparison site is named as the source. The next year the lender publishes the findings on an HTML page with dated, self-contained findings, a method section and a citation line, and asks the comparison site to credit and link the original. The share of coverage crediting the lender is the metric to watch, not rankings.
What I would change first
If I had one quarter, I would move the flagship study to an HTML findings page with a method section and citation line, build proper bio pages for the three experts the company most wants quoted, and start logging how journalists found us. That covers both the traditional search path and the assistant path, and in a year it gives you evidence about whether the discovery shift is real for your sector instead of an assumption.
Need help with this on your own site?
Niraj Raut works with businesses in Nepal, Australia, the UK, Europe and the US on technical SEO, ecommerce SEO, local SEO and AI search.
Add a reply Cancel reply
Replies are open to approved members. Log in or apply to join.