Discussion starter: an AI search 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.
- AI Search
Do AI answers favour content with first-hand signals like original photos, test data and named authors over expert summaries? Any tests?
Disclosure: Niraj Raut, who posted this answer, runs the SEO consultancy linked at the end of it.
1 reply
-
adminAdmin
There is no controlled public test that I can point to showing AI answers prefer first-hand signals over expert summaries, but the mechanics suggest some first-hand signals help more than others. Original test data and specific findings are likely to help, because they give an answer something unique to quote or cite. Named authors help attribution and credibility, especially where AI features rely on search systems that value demonstrated expertise. Original photos matter much less for text-based retrieval, although they matter for visual search and for readers. Google’s guidance on helpful content emphasises first-hand experience and expertise, and its AI features draw on its ranking systems, so first-hand content that ranks well is more likely to be cited. The effect probably runs through usefulness and ranking rather than through a special preference for first-hand markers.
Which signals plausibly matter, and why
Signal Likely effect on AI citations Reason Original test data and measurements Positive Unique, specific facts that answers can quote Specific first-hand observations Positive Distinctive passages; answer questions summaries cannot Named, credentialed authors Modestly positive Attribution and trust, especially on expertise-heavy topics Original photos Small for text answers; larger for visual search Text retrieval does not read images; multimodal systems may Expert summaries of existing information Neutral to weak Interchangeable with many other summaries Why specificity is the common factor
An expert summary that restates what is widely known competes with dozens of similar pages, and an assistant can cite any of them, or none. A page that says “across four weeks of testing, the battery lasted 31 hours on mixed use, against the claimed 40” contains a fact nobody else has. If a user asks about the battery life, that page has a passage nothing else can replace. First-hand work tends to produce that kind of specific, attributable information, which is probably why it seems to do well.
How to test it on your own content
- Select pairs of pages on similar topics with similar rankings: one built on first-hand testing, one an expert summary.
- Define prompts that each page could answer.
- Run prompts repeatedly across several assistants and record citations.
- Compare citation rates, controlling for rankings, since pages that rank better are cited more.
- Then add first-hand data to some summary pages and see whether their citation rates change.
Hypothetical example:
A home technology publisher has two kinds of review: hands-on reviews with measured results and photos, and roundup pages summarising specifications. Prompt tracking over three months shows hands-on reviews cited far more often for questions about real-world performance, such as battery life, noise and set-up time, while roundups are cited occasionally for specification questions, often alongside manufacturer pages. When the publisher adds its own measurements to five roundups, those pages begin to be cited for performance questions too. The result points to specific data, not the review format, as the factor that matters.
Watch for this:
Adding the appearance of first-hand experience, such as generic stock photos presented as testing or phrases like “in our experience” without any real testing, does not create the specific information that makes first-hand content useful, and it misleads readers. Only claim testing you have done.
Named authors: what they add and what they do not
Naming a qualified author tells readers and search systems who stands behind the content, and on expertise-heavy topics that supports credibility. It does not, on its own, make generic content more useful. An expert summary with a named author is still an expert summary. Where names seem to help most is when the author’s expertise is visible elsewhere: publications, professional profiles, prior coverage. That makes the attribution meaningful. For AI answers that name sources, an identifiable author also makes it easier to attribute a finding to a person and publication, which is a form of visibility in its own right.
What to invest in
For topics where experience changes the answer, such as products, services, tools, travel and procedures, invest in real testing and first-hand observation, and write findings as specific, quotable statements with numbers and context. For topics where the answer is settled knowledge, clarity and accuracy matter more than first-hand markers. Name authors with relevant expertise either way, and show how conclusions were reached.
Invest in first-hand testing where it produces specific findings, state those findings clearly with numbers and context, name qualified authors, and treat the citation effect as a consequence of more useful, distinctive content rather than a response to first-hand signals as such.
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.