[AI Search Webinar] AI Visibility You Can Measure: AI Search Reporting For Marketers

AI Search is already reshaping how brands are discovered, but how do you track that visibility and connect it to real business impact?

In this webinar, Scrunch AI CEO Chris Andrew and SUSO’s Partner Growth Manager Cara Corbett unpack how marketers can measure, report, and communicate performance across AI models like ChatGPT, Perplexity, and Google’s AI Overviews.

They explore:

➡ Frameworks for measuring AI Search visibility

➡ The tools enabling accurate AI Search reporting

➡ How to communicate AI Search success in a wider marketing context

Scroll down to watch the recording.

Click here to access the presentation slides.

Webinar Transcript

00:00:00.000 –> 00:00:14.744 Hi, everyone. I’m Cara Corbett from SUSO Digital, an SEO firm that provides free SEO and AI search training to agencies to help them thrive in the new era of AI search.

00:00:15.660 –> 00:00:24.464 Welcome to our webinar, “AI Visibility You Can Measure,” which is the third and final part of our AI Search series.

00:00:25.244 –> 00:00:49.617 In the first webinar, we explored the AI search opportunity and how agencies can capitalize on it. The second was tactical, focusing on driving visibility through structured content and technical accessibility.

00:00:50.058 –> 00:01:06.329 Today, we’re focusing on the next step: measurement and reporting. We’ll cover how to track brand visibility, understand its quality, and demonstrate that value to clients.

00:01:06.629 –> 00:01:21.814 I’m joined by Chris Andrew, CEO of Scrunch. I can vouch for his candid perspective on what is actually possible in measurement today.

00:01:22.234 –> 00:02:21.215 Yeah, thank you, Cara. I’m Chris Andrew, CEO of Scrunch. We’ve built a platform to help marketers manage the shift to an AI-first customer journey, understand how they show up in AI search, and better interact with AI agents and crawlers.

00:02:21.695 –> 00:02:34.003 As always, we’ll have time for questions at the end. Feel free to use the chat bot.

00:02:35.038 –> 00:03:22.613 Today, we’ll explore how brands appear in platforms like ChatGPT, Perplexity, and Google’s AI Overviews. Then, we’ll dive into the actual KPIs for measuring AI search visibility and the methods and tools to track them, from manual testing to platforms like Scrunch and traditional tools.

00:03:23.774 –> 00:03:59.162 You know that traditional SEO success was measured by keyword rankings, organic traffic, and appearing in the top blue links, which led to traffic and leads.

00:03:59.962 –> 00:04:25.660 But with AI search tools, the game has changed. Success is no longer just about blue links; it’s about appearing in AI-generated answers. This is often called “zero-click search.”

00:04:26.520 –> 00:04:48.980 For example, on ChatGPT, visibility comes from a direct brand mention, like “HubSpot’s a popular choice,” which positions your brand as a recommended solution.

00:04:49.899 –> 00:05:13.333 On Perplexity, your brand is cited as a source in AI summaries. These citations position you as an authority and can hypothetically lead to website traffic.

00:05:14.514 –> 00:05:43.980 In Google’s AI Overviews, visibility means appearing in featured snippets or summarized answers, which gives exposure at the top of the user journey and takes up a lot of real estate.

00:05:44.821 –> 00:06:09.934 And in Google’s AI mode, your brand can appear in real-time AI summaries alongside traditional results. Success looks different depending on the platform.

00:06:10.454 –> 00:07:00.710 AI tools don’t just provide a ranked list; they synthesize answers from multiple sources, going deeper than the first page. Even ranking number one in Google doesn’t guarantee an AI mention. A study found brands on Google’s first page were mentioned in ChatGPT only 62% of the time. It’s now about being part of the answer.

00:07:01.739 –> 00:07:16.183 So, what are the key metrics we should track? I’ll pass it to Chris for a disclaimer on what’s possible.

00:07:17.254 –> 00:08:07.930 Yeah, thanks. We encourage companies to think practically about what you can irrefutably measure. First, dig into your web logs. Understand the frequency with which AI agents and crawlers are visiting your website. This is a leading indicator of a new customer journey.

00:08:08.579 –> 00:08:34.809 Next, you can track referral traffic and conversions. In Google Analytics, you can see how many visitors you’re getting referred from GPT, Perplexity, etc., and understand their conversion rate.

00:08:35.189 –> 00:09:13.707 In the middle, you can measure sentiment and perception towards your brand. This can be done manually in an incognito browser by asking key questions, or at scale with a tool.

00:09:14.438 –> 00:09:50.462 Yeah, and some metrics are a little more “hazy” because LLMs don’t give access to their search data.

00:09:51.403 –> 00:10:09.520 One of these is Share of Voice, which evaluates your brand’s presence relative to competitors across AI platforms.

00:10:10.261 –> 00:10:26.823 Brand Mentions helps you track if your brand is mentioned in AI-generated responses, assessing direct visibility which can influence user perception and trust.

00:10:27.263 –> 00:10:52.204 And Brand Citations is when your brand is cited as a source. These contribute to your brand’s authority and can sometimes lead to traffic.

00:10:54.196 –> 00:11:48.650 There are questions about who should own these KPIs. Is it PR, content, or SEO? The truth is, it’s the intersection of all three. We need collaboration between these teams to feed AI systems a unified and credible brand story.

00:11:49.489 –> 00:12:12.776 So, how do we measure these? We’ll show you methods and platforms to monitor visibility and benchmark against competitors. Together, these methods form a complete view.

00:12:13.639 –> 00:13:08.185 One of the simplest ways is manual prompt testing. You simulate user questions in ChatGPT, Perplexity, etc., and see if your brand appears. You can test prompts for brand perception (e.g., “What is [brand name] known for?”) or competitive positioning. Use an incognito browser for unbiased answers.

00:13:08.645 –> 00:14:01.627 If the AI isn’t giving the right answers, that’s a signal to take action. This could include updating website content, creating structured FAQs or blog posts, or engaging in a review strategy to ensure consistent messaging.

00:14:03.309 –> 00:14:53.906 Obviously, AI tools don’t give you insight into actual user prompts, but you can find user intent from sources like Google’s “People Also Ask” feature and Google Search Console, which reveals the actual search queries people use to find your site.

00:14:55.046 –> 00:15:26.978 Your sales and support teams are also a wealth of knowledge about common questions. You can also check your site’s search bar data and use social listening on sites like Reddit.

00:15:27.558 –> 00:15:58.212 Use these insights to inform your strategy. You’ll want to run weekly or monthly tests to track how your visibility changes, either manually or with scaled tools like Scrunch.

00:15:58.712 –> 00:16:10.756 So now I’m gonna hand it off to Chris for a brief demo of how his platform helps measure AI search performance.

00:16:12.544 –> 00:16:46.918 Awesome. Thank you. Agencies are in a great position to capitalize on this shift. The Scrunch platform is partner-first, designed for an agency to be completely self-serve.

00:16:47.178 –> 00:17:53.041 The starting point is visibility into what AI agents are doing on your website. This screen shows the frequency of visits from agents like GPT, Perplexity, and Google. We can see they are either retrieving content live, training future models, or indexing the web.

00:17:53.821 –> 00:18:32.682 We can see the top pages they’re visiting and a live feed of this activity. Irrefutably, AI agents are on your site today answering questions on behalf of humans, and you can measure this.

00:18:33.243 –> 00:19:42.097 The next question is, what are people asking? Using Disney+ as an example, our “AI Search Trends” feature aggregates third-party data to give a directional understanding of the types of questions consumers are asking, grouped by topic.

00:19:43.340 –> 00:20:48.178 Next, we audit what happens when an AI agent visits the site. A visual site like Disney+ might have low “content quality” for an agent, which is looking for language, not JavaScript or multimedia. We automate these audits.

00:20:49.381 –> 00:22:16.389 Then we move into brand presence, citation rate, and sentiment. This dashboard shows how frequently Disney+ appears in results compared to competitors like HBO and Netflix, with deep filtering available.

00:22:16.710 –> 00:22:26.760 For each prompt, you can click in to see the actual day-by-day results from the model and see what content is being cited to answer the question.

00:22:27.100 –> 00:23:42.007 The customer journey is fundamentally changing. Humans are outsourcing browsing. We also highlight citation rate—which websites are being cited most frequently for the questions you care about. This is hugely actionable for PR and content teams.

00:23:42.407 –> 00:24:08.598 Today’s demo focused on visibility, but we also have an optimization component. I’ll pass it back to Cara.

00:24:08.678 –> 00:24:58.179 Thanks, Chris. Other SEO tools are adapting, too. Ahrefs has introduced “Brand Radar” to monitor mentions, citations, and share of voice. Even its basic Site Explorer gives a high-level snapshot of AI Overviews and ChatGPT citations.

00:24:58.939 –> 00:25:47.250 SEMrush offers an “AI Visibility Checker” that also tracks mentions, citations, and analyzes sentiment (positive, negative, or neutral). These tools can also leverage traditional SEO data to find content gaps.

00:25:47.790 –> 00:26:21.595 But remember, because LLMs don’t provide real search data, much of what we measure from prompts is synthetic. The most tangible and accurate insight you can get today is by tracking referral traffic and conversions.

00:26:22.862 –> 00:27:47.953 I’ll show you how to do this in GA4. Go to Reports > Acquisition > Traffic acquisition. Add a filter for “Session source” and set the match type to “matches regex.” We’ll share the regex code snippet that includes all the AI tools. Then, adjust your date range.

00:27:48.794 –> 00:28:17.561 Right now, it’s grouped by channel. Change the primary dimension to “Session source.” Now you can see the breakdown of traffic coming from ChatGPT, Gemini, Perplexity, and others.

00:28:18.062 –> 00:29:31.914 To see which pages they’re sending traffic to, go to Engagement > Landing page. Apply the same regex filter as before. Then, add a secondary dimension of “Session source.” Now you can see exactly which AI tool sent traffic to which page on your site.

00:29:33.053 –> 00:30:47.106 Before Q&A, I want to mention our free Partner Club. It gives agencies and consultants access to free personal workshops on SEO and AI, a 24/7 help desk with our SEO specialists, and free SEO and visibility reports, which are analyzed by our team.

00:30:48.687 –> 00:31:29.640 That brings us to our Q&A.

00:31:30.040 –> 00:32:23.370 I’ll add one thing: this shift is a huge opportunity for the SEO and agency space. SEO specialists have been “marketing to technology”—the Google bot—for decades. This is just an expansion of that, with new agents and crawlers coming to your site.

00:32:24.910 –> 00:33:03.691 We have a question from Haley: “Do you think a company needs to be a certain size?” Larger companies will definitely have a leg up due to brand awareness and dedicated teams, but that doesn’t mean smaller companies can’t employ tactics to get results.

00:33:03.691 –> 00:34:51.004 Haley also asks how this differs for B2B. I think B2B is actually more important. People are using AI search for highly considered purchases and asking very thought-out, deep questions, like comparing vendors or finding niche parts. For small teams, just start with the manual spreadsheet approach.

00:34:51.588 –> 00:34:59.250 Yeah, and B2C is about to explode, with ChatGPT announcing integrations with Shopify, Etsy, and PayPal.

00:34:59.270 –> 00:36:09.293 Amelia asks: “How does schema code work?” Schema markup is really important. It’s code that signposts your content for AI. For example, FAQ schema helps AI pull direct questions and answers, and it also enhances your visibility on traditional search results.

00:36:10.153 –> 00:36:58.180 Selim asks: “When you say incognito mode, is the temporary chat on ChatGPT considered incognito?” That’s a good question. I don’t think it is. My understanding is it’s still using your context but just not saving the chat. I would just log out completely to be safe.

00:36:59.100 –> 00:37:32.732 Hojoon asks: “Does your platform provide recommendations?” Yes, it does. We provide recommendations on gaps to fill, and we also have an “agent experience platform” that transforms existing site experiences into ones optimized for AI consumption.

00:37:33.973 –> 00:38:39.081 Another question: “In Scrunch, are the questions shown only the questions being searched or the top questions?” Nobody has access to the “top questions” except the AI platforms themselves. What we have is “AI search trends,” which is a representative sample of real questions consumers are asking. You can also add your own custom prompts to monitor.

00:38:41.289 –> 00:39:05.045 Okay, I think we’ll call it there. We’ll share this deck, and our emails are linked. If you join the partner club, you get access to the help desk, so feel free to send us questions.

00:39:05.345 –> 00:39:11.949 Thanks so much, Chris, for joining. And we’ll see you all on the next webinar. See you, everybody.