Why did our AI mentions jump after we added personal experience?

We saw our AI mentions jump 40% in a week when we stopped writing polished generic posts and started putting our own lived experience into them. That change brought more impressions, more users, and better intent, because people asking ChatGPT or Perplexity what to use are already close to a buying decision.

What are ai trust signals?

AI trust signals are the bits of proof that make ChatGPT, Claude, Perplexity, and Google comfortable enough to mention you by name. Across sources, the pattern is boring and consistent: clear identity, evidence, named authors, citations, and pages a machine can read without guessing. Trust Signals describes them as the credibility layer AI uses when buyers ask for recommendations, and Semrush frames the same thing as identity, evidence, and technical health.

That matters because AI doesn't behave like ten blue links. A buyer can compare ten search results and make up their own mind. An AI answer often gives them a short list, sometimes just one or two names. AITrustSignals.com says it plainly: Google gave buyers ten links, AI gives them one answer.

If you run a small SaaS, don't hear this as "go produce more content." Hear it as: give the model enough proof to repeat your claims safely. NoGood breaks trust signals into three buckets: entity identity, evidence and citations, and technical clarity. That's the cleanest simple model I've seen.

The practical point is simple. AI won't trust what it can't verify. If your site says you're the fastest, easiest, or best, but there's no named person behind the claim, no source, no customer proof, and no clear explanation of what your product actually does, you gave the model ad copy, not evidence.

Why did our AI mentions jump after we added personal experience?

First-hand experience makes your claims easier for AI to trust because it gives the model something concrete to grab onto. When we added our own operator experience to blog posts, our AI mentions jumped 40% within a week. Not because the prose got prettier. Because the posts stopped sounding like every other SEO page and started showing how we knew what we knew.

That lines up with what the sources say. SEOptimer ties AI trust signals back to experience, expertise, authority, and trust. IMPACT says AI looks at the depth and structure of your content, plus how transparent you are about pricing and operations. And the SIGI paper found that methodology sections scored 8.5, while named authorship with credentials and source citations each scored 7.5 in a model self-report about citation decisions.

Personal experience does three useful things at once:

• It shows there is a real person behind the claim
• It gives the model traceable context, not vague marketing language
• It creates details that are hard to fake and easy to quote

A founder saying, "we changed this onboarding step and activation moved," is stronger than a generic post about onboarding best practices. A builder showing the mistake they made, the fix they shipped, and the result they saw is easier to trust than a page stitched together from other people's summaries.

That's why I'd fix the evidence inside your existing content before I wrote ten new posts. Volume is easy. Proof is what gets repeated.

Why doesn't ranking in Google guarantee AI mentions?

AI systems use a different filter before they ever get to the part where they generate an answer. NoGood says ranking well in traditional search doesn't automatically put you in front of users on AI platforms, because AI answers pull from a different source mix and apply different standards. AuthorityTech goes further and says trust signals gate two stages of the retrieval pipeline, including whether your content even makes it into the candidate set.

That's why some weaker products get mentioned more than better ones. They left a readable trail. You left a landing page, a changelog, and some tweets.

One useful number from NoGood is that earned third-party coverage drives roughly 72% of AI citations. Even if that number shifts by platform, the direction is obvious: the web around your site matters, not just the site itself.

The same source also says platforms weight signals differently:

• ChatGPT leans on training data and canonical entities
• Claude rewards primary sources
• Perplexity favors freshness and community
• Grok depends on X presence
• Gemini and Google AI Mode track closest to search

So yes, SEO still helps. But SEO alone doesn't solve the citation problem. If your product is hard to classify, your claims are unsupported, and your site has no visible humans attached to it, AI has no safe reason to say your name.

Which trust signals should a solo SaaS fix first?

The fastest wins are usually the pages and details you skipped because they felt boring. AITrustSignals.com literally shows missing Terms of Service as a quick win, with an example line from ChatGPT saying it couldn't locate that page for a regulated service. That's not glamorous. It is exactly the kind of gap that keeps AI from trusting you.

For a tiny SaaS, I'd fix these first:

1. Named authors on articles and docs. SIGI found named authorship with credentials scored 7.5.
2. A methodology or "how we know this" section on opinionated or data-backed posts. SIGI scored methodology highest at 8.5.
3. Source citations inside your content. Same paper, same 7.5 score.
4. Clear pricing, terms, and privacy pages. IMPACT says transparency about pricing and operations is part of what AI evaluates.
5. Customer proof with names, use cases, and outcomes. AI can do more with a real story than with "loved by modern teams."
6. A plain-English product page that says what the tool does, who it's for, and when not to use it.
7. Basic company identity like founder names, company details, contact info, and consistent descriptions everywhere.

Most founders overvalue net-new articles and undervalue these trust pages. That's backwards. If the foundation is thin, more content just spreads the problem across more URLs.

Why do named authors, citations, and methodology matter so much?

Verifiability beats polish because AI has to decide whether it can repeat your claim without embarrassing itself. The SIGI research paper is useful here because it gets very specific. In its model self-report, methodology sections scored 8.5, named authorship with credentials scored 7.5, and source citations scored 7.5. Publication dates scored 7.0 and last-updated timestamps scored 6.5.

That stack tells you what the model is looking for. Not vibes. Not content marketing theater. Proof it can inspect.

A methodology section doesn't need to sound academic. For a SaaS blog, it can be as simple as:

• what data you looked at
• what time period you used
• what you changed
• what happened after
• what the limits were

A named author doesn't need to be a celebrity. It just needs to be a real person with a reason to know the topic. If the founder wrote the post because they shipped the feature, say that. If your engineer ran the test, put their name on it.

And citations matter because unsupported claims force the model to choose between silence and parroting marketing copy. Good AI systems lean toward silence. That's why adding your own experience helps so much when it's paired with receipts. The story gets attention. The proof gets cited.

How do you make your content easier for AI to cite?

Clean, extractable pages beat clever writing because retrieval systems need to parse your point fast. AuthorityTech includes content extractability and technical accessibility as core dimensions, and says a source missing one of those can get filtered out before the language model sees it. Semrush says the same in plainer language: identity, evidence, and technical health decide whether your brand looks credible enough to cite.

For most SaaS sites, that means:

• one page, one job
• obvious headings that say what the section answers
• direct claims followed by proof
• pricing in plain text, not hidden behind demos
• docs that explain inputs, outputs, limits, and setup
• FAQs with full-sentence answers
• dates, updates, and author names visible on-page

You don't need to write for robots. You need to stop making humans and robots guess. A page that says, "AI meeting notes for recruiters" is better than a vague headline about transforming conversations. A case study with a named customer, workflow, and outcome is better than a logo wall. A founder note explaining why you built a feature is better than a paragraph of filler adjectives.

We saw that first-hand. Once we started writing posts with our own examples, our own decisions, and our own outcomes, AI mentions moved fast. The machine finally had something specific to retrieve.

What should you stop doing if you want more AI mentions?

Generic blog volume stops helping the moment every claim sounds like it came from the same prompt. AI systems already have enough bland summaries. What they need from you is evidence they can't get from everyone else.

Trust Signals says brands often think they did everything right and still don't show up in AI answers. The missing piece is usually not effort. It's that nobody turned their credibility into signals the machine can evaluate. NoGood says entity consistency and schema validation are often the fastest wins because the groundwork already exists. That's a good reminder that the fix is often operational, not creative.

So stop publishing pages that:

• make a big claim with no source
• hide the author
• skip pricing and policy details
• copy what every competitor already said
• bury the useful part under a fluffy intro

If you only change one thing this week, add first-hand experience to the pages that already rank or already get crawled. Show what you did. Name who did it. Cite what supports it. Put the boring trust pages live.

That's the pattern behind most AI mentions I've seen, including our own. The lift didn't come from more content. It came from making our content believable enough to repeat.