AI Visibility Measurement: What To Track & What To Ignore
Webinar New AI Search & SEO KPIs: 4 Real Signals AI mentions and citations are benchmarks, not decisions. Get 4 traffic-predictive signals drawn from real bot data across hundreds of sites. Rundown Inside AI Max, PMax & Smart Bidding What to measure, which signals to prioritize, and how to guide automation before AI Max becomes the default. Rundown Why AI volume alone can't deliver personalization Your content architecture may be the real bottleneck. Get the Rundown on what needs to change in your content ops. Webinar New AI Search & SEO KPIs: 4 Real Signals AI mentions and citations are benchmarks, not decisions. Get 4 traffic-predictive signals drawn from real bot data across hundreds of sites. Webinar How Freshpet Earned AI's Trust The GEO playbook behind Freshpet's AI Overview citations, plus checks to run on your own brand. Free, live Aug 20. 🔥[Live 8/12 with Loren Baker] Ecommerce SEO: Own your "brand +promo code" search. Track prompts, cross-model visibility, and self-reported attribution instead of citations and referral traffic, where 80-90% of AI leads get mislabeled organic. I have dozens of conversations per week with folks in growth and marketing, ranging from directors, VPs of marketing, and CMOs to SEOs in the nitty-gritty day-to-day. Many of my conversations involve measurement. This is an increasingly challenging topic as traditional SEO metrics are breaking down with the advent of AI answers. Much of the conversation is spent debunking misconceptions and misguided advice operators see on social media. It’s not an easy conversation, but it’s important. This means telling people to avoid many of the things they see people promote on social media. It’s not an easy conversation, but it’s the right conversation. My goal here is to clear up the confusion so you can tie your AI visibility efforts to business outcomes. These are a mix of leading and lagging indicators that you have varying degrees of control over. I’ll make the case for why each one matters, then we’ll cover how to influence them. This is the most obvious and most important decision, because what you measure influences behavior. From dozens of conversations, it’s also what many people get wrong. It becomes the first domino in a chain of mismeasurement. None of the other metrics, like citation share, brand mentions, or visibility, matter if you’re tracking the wrong set of prompts. Most AI visibility tools like Profound, Peec, and AirOps will automatically recommend prompts to track, but these are rarely what you should focus on. I haven’t confirmed this, but from what I can tell, they analyze your website and map existing pages back to prompts. They assume the pages already on your site are the ones that should be cited or visible in LLM outputs. That might be true. But most companies we speak to say, “We aren’t appearing for the prompts we want to show up for,” which tells me they don’t have the right strategy and thus haven’t published the right pages. You can see how, if you’re tracking the wrong prompts, you’ll measure visibility for the wrong things. From our perspective across dozens of clients, ChatGPT is the most commonly used LLM, but it’s worth tracking how often your brand shows up for your target prompts across ChatGPT, Gemini, Google AI Mode, Perplexity, Claude, and Copilot as well. This is important because ChatGPT’s user base skews consumer, while Claude’s user base skews business and enterprise. If you’re a B2B business, Claude has fewer users, but those users are using Claude at work, which is the context that matters. Google has also made AI mode more prominent, and with Chrome having dominant market share, many people will see AI mode outputs by default. Tracking each model separately lets you see whether a change in visibility or traffic is isolated to one model or happening across all of them. There are also significant differences in how each model searches for and presents information. For example, Claude’s use of Brave versus ChatGPT’s use of Bing; they have their different methodologies for selecting sources; and each model has preferred sources through media partnerships. For most of the last decade, marketers leaned on clickstream analytics and UTM parameters to tell them where leads came from. That paradigm has been breaking down for a while, and it now shows an even smaller part of the picture. LLMs are zero-click by design. When someone asks ChatGPT, Gemini, or Claude a question, the answer is provided in the chat. They don’t click through. They research in the conversation, then they might do a Google search for your brand directly or type your URL straight in. Your analytics and CRM platforms will log that traffic as “Organic” or “Direct,” but the reality is that an LLM is what led to the person going to your website. Self-reported attribution is the simplest yet highest-signal way to get more of the story. We recommend simply asking people, “How did you hear about us?” There are well-known and documented flaws to self-reported attribution (mostly human memory and salience of touchpoints), so it won’t be perfect, but it’s the buyer telling you, in their own words, how they found you. If you capture that information through a field on your lead form, you can then track those lead sources in your CRM to pipeline and closed revenue. That’s the ultimate success metric for a marketing channel. One client that added it discovered ~5% of registrations were coming from ChatGPT despite doing zero work around AI visibility. Below is a chart from our internal dashboard. These are the number of people who filled out our consultation form and stated that they found us through an LLM, split by the HubSpot-tagged source. Where possible, HubSpot automatically tags the source as “AI Referrals” (red). We found that 80-90% of leads that came in via an AI platform were incorrectly tagged as “organic” or “direct.” These metrics tend to be related but not the ultimate goal, and often not what you have control over. This is the strategic piece that people mistakenly view as the success metric. Off-page sources dominate citations at every funnel stage. Our research on citation sources found that, even for branded or bottom-funnel queries, 48% of sources were earned media, 30% were commercial content from other sites, and only 22% were from the brand’s own website. So you should be measuring what percentage of attainable mentions you currently occupy across third-party surfaces. If it’s 1%, there’s massive headroom.
Source: Search Engine Journal
This article has been carefully curated and reformatted for educational and informational purposes. Full credit goes to the original publisher.
📚 Visit more helpful articles on Joab Peters Blog
No comments