Latest / Found in AI: AI Search Visibility, SEO, & GEO / What is an AI Visibility Audit? (And Do You Need One?)

Transcript

What is an AI Visibility Audit? (And Do You Need One?) · Found in AI: AI Search Visibility, SEO, & GEO

• 0:41
  (Transcribed by TurboScribe.ai. Go Unlimited to remove this message.) Okay, if it's not immediately clear already, I am a data nerd.
• 0:46
  I can't help it, I just really like numbers and patterns.
• 0:50
  And so, I say a lot to say, I have been digging through my Google search console data lately and one thing keeps popping out at me.
• 1:00
  A lot of brands must be asking the same question and it's all about AI search visibility audits.
• 1:06
  I'm guessing the question is, do we need an AI search visibility audit or how do we even conduct one?
• 1:12
  Hey, I'm Cassie Clark, a fractional content strategist, AI search optimization expert, sort of, and the host of Founding AI, the show dedicated to helping marketers and founders learn AI search and GEO strategies.
• 1:27
  Since this keeps reappearing in my Google search console data, I thought today would be a great time to really just break it down, particularly what an AI visibility audit actually is, why brands are asking for it now, and how I think about audits through my freshness, structure, and authority framework, the FSA framework.
• 1:45
  Okay, let's dig into all of this.
• 1:57
  Okay, before we really dig into this, let me give a general PSA.
• 2:02
  I am dealing with a case of the stumblies again today.
• 2:06
  It's either I've not had enough coffee or the brain fog is working overtime or it's a combination of both.
• 2:12
  That's what I'm thinking it is right now, but I appreciate you sticking around.
• 2:16
  I will get it out eventually.
• 2:18
  So, let's really just start with what an AI visibility audit actually is.
• 2:24
  At a high level, it looks at how AI engines like ChatGPT, Gemini, Perplexi, and now MetaAI interpret and recommend a brand when users ask real questions.
• 2:37
  Now, all of this matters because AI-driven discovery doesn't work like traditional search.
• 2:44
  An AI visibility audit is not about rankings.
• 2:47
  It's not a keyword report and it's definitely not a dashboard full of screenshots showing whether you appeared once or twice in a list.
• 2:56
  Screenshots will probably be included in your report, like in your audit findings, but it shouldn't just be all screenshots.
• 3:05
  An AI visibility audit helps brands understand how AI engines perceive it.
• 3:12
  It's really asking the questions, does the AI engine understand who you're for?
• 3:17
  Does it understand the problem you solve?
• 3:20
  And do you send strong signals, those strong confidence signals, off enough that the AI engine can recommend you when the question actually fits?
• 3:30
  Now, every audit I run follows the same framework.
• 3:33
  It's the FSA framework.
• 3:34
  I've talked about this in previous podcast episodes.
• 3:37
  It stands for Freshness, Structure, and Authority.
• 3:41
  If even one of these is slightly off, visibility tends to drop a little.
• 3:45
  Most teams don't really know which of these is causing the issue, and that's the gap that an AI visibility audit helps close.
• 3:53
  So, if you've been hanging around the Found in AI podcast for a while, again, you know I'm a nerd.
• 3:57
  I love data.
• 3:59
  I love numbers.
• 3:59
  I love staring at my Google Search Console reports, and honestly, it really bugs me that they don't update in real time, but I keep seeing AI visibility audit appear in that data.
• 4:12
  So, clearly, teams are feeling that something has changed, and they're trying to understand what is different before they actually go out and change something with their AI search optimization efforts.
• 4:23
  Now, AI search optimization is still very new, and measuring is kind of challenging.
• 4:30
  So, I say all that to say you should absolutely measure before you try to optimize, given the best tools that we have now.
• 4:38
  Hacking away at tactics without understanding what's broken, it usually just causes more frustration and wastes time.
• 4:44
  What's happening now is that we're seeing that these brands are noticing that their competitors are showing up in AI answers.
• 4:52
  They're also hearing prospects mention, hey, I found you in chat GPT, and maybe they weren't expecting that, and they're also starting to realize that visibility is changing before traffic numbers fully reflect that, although some teams are seeing the numbers already shifting.
• 5:08
  I get it.
• 5:09
  I absolutely get it.
• 5:11
  The instinct is to jump straight into, hey, we need more content.
• 5:15
  We should rewrite our pages.
• 5:17
  We need to do GEO now.
• 5:20
  I'm going to hold your hand when I say this, because most brands don't actually have a visibility problem, especially if you've been operating online for years.
• 5:28
  You already exist.
• 5:29
  You already have a presence.
• 5:32
  What you have is a diagnosis problem.
• 5:35
  That's why fixing freshness when the real issue is authority doesn't work, or tweaking your structure when relevance is the real gap is not very effective, or chasing visibility without understanding how these AI engines make decisions in the first place.
• 5:51
  That's why you need an AI visibility audit.
• 5:53
  That's why we need to measure things, again, in the best way that we can right now.
• 5:58
  That's why we need to understand what's going on before we make any changes to our strategies.
• 6:03
  The goal of an audit isn't really to reverse engineer AI engines or chase individual answers.
• 6:10
  The goal is to identify those repeatable patterns in what AI systems tend to favor, and then go in and change those patterns if we need to.
• 6:19
  Now, back in December, I conducted an AI visibility audit for a brand, and I wrote a case study about it.
• 6:26
  I want to go through what we actually looked for inside of this audit, so you get an idea of what your audit should look like as well.
• 6:35
  For this particular brand, I tested a range of natural language prompts, from broad discovery questions to very specific use case driven ones.
• 6:44
  For each prompt, I looked at which competitors appeared in those AI generated answers, how they were described, and which sources were cited.
• 6:54
  Then, I compared the brand's positioning and messaging to their competitors that showed up more often, not just on their website, but across blogs, and social content, and other public mentions.
• 7:08
  At this point in what we know about AI search visibility, we know that cross-channel and third-party mentions do play a big role in building into the authority, so I paid close attention to where this brand might be slightly underrepresented.
• 7:24
  What stood out very quickly was that clear messaging mattered even more than just traditional authority metrics like domain authority.
• 7:32
  In traditional SEO, we know that the higher your domain authority is, the better off you will be.
• 7:39
  It's kind of a strong predictor of visibility within those traditional search engines, but in AI generated answers, the pattern doesn't always hold.
• 7:49
  AI engines, they regularly cited the competitors, even though they had lower domain authority, when their positioning made it easier to understand who the product was for and when it should be recommended.
• 8:03
  This particular brand is a presentation maker.
• 8:07
  We also noticed that similar software, those other presentation makers, appearing most frequently in the answers were extremely explicit about their use cases.
• 8:19
  They had those use case pages clearly listed on their website.
• 8:25
  Hey, this is who it's for.
• 8:26
  Here's how they use it, and they also created that content and shared it across other channels, so instead of trying to appeal to everyone who makes presentations, they really anchored themselves to specific audiences or scenarios.
• 8:40
  This brand's positioning makes sense to human readers.
• 8:44
  If you go look at their website, it's a great website, you're going to see that, oh, this is for anyone who makes a presentation.
• 8:51
  But AI engines, they don't really like that just kind of wide positioning.
• 8:58
  Consistent, it makes it harder for consistent interpretation over time.
• 9:03
  So, when we looked at those cited sources and the brands that kept getting mentioned, we noticed that they were very clear.
• 9:12
  We had best presentation maker for students, best presentation for marketers, presentation maker for marketers, best presentation maker for legal teams presenting cases.
• 9:23
  Very, very specific use cases.
• 9:26
  And when that messaging is reinforced consistently across channels, those AI engines really learn exactly what a product does and who it helps.
• 9:35
  Another key finding from this audit was that the brand had relatively few off-site signals.
• 9:42
  What I mean by that is their competitors, especially the ones that appeared more frequently, had a strong presence beyond just their websites.
• 9:51
  They were very consistent with their language across blogs and social channels and YouTube or wherever else they were posting.
• 9:59
  In that consistency, it really helps those AI engines form stronger associations between the brand and specific presentation needs.
• 10:07
  These competitors were very active on YouTube, Instagram, LinkedIn, TikTok, Reddit, wherever their audiences actually are.
• 10:18
  And those tend to be the same places that AI engines pull those signals from.
• 10:23
  So, cross-channel activity directly strengthens that authority layer for the FSA framework.
• 10:29
  And it's really something that your audit should take account of.
• 10:34
  We also looked at freshness, too, in the audit.
• 10:37
  Freshness is not just about publishing more.
• 10:40
  It's really about the same positioning showing up repeatedly in those current conversations.
• 10:47
  Volume matters far less than consistency, and you really need to be specific in your messaging when you're consistently posting over time.
• 10:55
  I keep coming back to this because it's so important.
• 10:59
  Most AI visibility problems are actually authority problems disguised as content problems.
• 11:04
  You have to be very clear about who you serve and why, and then reinforce that everywhere in your content.
• 11:12
  So, when you're looking for someone to run an audit for your brand, or if you're doing it yourself, you really should expect some baseline numbers.
• 11:22
  Things like AI share of voice and a few other visibility metrics.
• 11:26
  I have another episode where I talk with Christina Fonse about what metrics that we should be tracking.
• 11:32
  I'll link that in the show notes.
• 11:34
  Measurement is still absolutely evolving.
• 11:37
  While there are plenty of AI tools on the market, the dust is still really settling about all of this.
• 11:43
  So, I personally, and again, it's because I'm a nerd.
• 11:47
  I prefer manual tracking for now.
• 11:49
  I've noticed that the more complex your prompts that you're tracking are, the more complex it becomes.
• 11:58
  So, you really need to only monitor your money prompts.
• 12:01
  I'm calling those money prompts.
• 12:03
  It's the questions where your brand should show up to influence those purchasing decisions.
• 12:08
  A good audit, it goes beyond those metrics though.
• 12:11
  It should help you understand how AI engines associate your brand with those specific use cases, so you show up when the question actually fits.
• 12:20
  Not just randomly, not inconsistently.
• 12:24
  Internally, a successful audit, it helps create focus for your team.
• 12:30
  It helps teams align around the messages that matter.
• 12:33
  It clarifies priorities and which channels you should focus on.
• 12:38
  It also defines what freshness, structure, and authority actually look like for your brand.
• 12:43
  Instead of working in disarray, that alignment allows those signals to reinforce each other.
• 12:50
  Now, over time, when everything's working as it should, and those audit insights are applied consistently, freshness, structure, and authority, it begins to compound.
• 13:01
  Those AI engines gain confidence.
• 13:03
  Those recommendations become more predictable.
• 13:06
  Invisibility becomes something that you can build intentionally.
• 13:09
  That's what a good audit does.
• 13:11
  It shows you where you are now, and it gives you the roadmap to increase AI's share of voice and visibility over time.
• 13:17
  As a fractional content strategist, I do offer AI visibility audits.
• 13:22
  Those audits really answer one core question.
• 13:24
  What's actually blocking your visibility and AI answers?
• 13:28
  Is it freshness?
• 13:29
  Is it structure?
• 13:30
  Is it authority?
• 13:31
  Now, when brands come to me for an audit, this isn't really a commitment for ongoing work.
• 13:35
  Some teams take those insights and fix things internally.
• 13:39
  Others ask for help putting everything in place for an actual strategy.
• 13:43
  Either way, an audit gives you clarity on where your brand stands and how to move forward.
• 13:48
  Once you know whether the issue is freshness, or structure, or authority, everything else in your AI search or GEO strategy, whatever we decide to call it, everything else becomes easier.
• 14:00
  That's it for this episode.
• 14:02
  If you want resources on how to run a baseline audit yourself, or if you'd like me to do one for you, I'll leave all of that in the show notes.
• 14:09
  Okay, now, a couple of things before I go.
• 14:12
  I can't end this episode without mentioning that OpenAI launched its ads program yesterday.
• 14:18
  I'll be diving into that more on Thursday.
• 14:22
  I haven't had time to fully dig into it yet.
• 14:24
  It's Tuesday morning currently.
• 14:26
  My brain was tired last night, but we'll talk about what Tent2BT ads actually look like now that the rollout is here.
• 14:33
  A lot of us have been waiting for this since it was announced last month, and honestly, I'm kind of surprised that it was so quickly.
• 14:39
  Anyway, also, as always, I know if you don't ask, but I'd love for you to subscribe, leave a rating, and this just really helps more marketers looking for answers find it a bit easier.
• 14:50
  I am forever grateful for each of you.
• 14:53
  Thank you for listening.
• 14:54
  I will see you in the next episode.
• 14:56
  Until then, stay visible.