As LLM usage exploded and the news around AI Overviews became a hot topic in 2025, our SEO-first agency faced the same crisis as others in 2025; AEO is brand new, how do we know how to optimize for visibility in LLMs? And how do we track and report these results to clients?
Every agency I talk to right now is navigating the same uncomfortable conversation. Clients are seeing ChatGPT and Perplexity traffic showing up in their GA4 reports, they’re reading headlines about AI eating Google’s lunch, and they want to know: what’s the plan?
Most agencies don’t have a good answer yet. They’re still sending ranking reports built for a search landscape that’s quickly becoming obsolete, focused on positions 1 through 10 on a Google SERP. Clean, linear, comfortable. And I get it. Nobody wants the game to change when they’ve gotten good at playing it.
But here’s the reality: if you’re running an agency and you can’t report on AI visibility yet, you’re already behind. The good news is it’s not as complicated as it sounds, and the tools have caught up to the need. I’ve been deep in this at EWR Digital for a while now, and I want to walk you through how we’re handling AEO tracking and reporting for our clients; what we measure, how we report it, and the tools that make it scalable.
Why LLM Visibility Percentage Is a Better Metric for AEO
I was recently talking with Connor Kimball over at Cairrot, and he introduced a concept that really crystallized something I’d been thinking about: search impression share. Once he explained how Cairrot frames it, it clicked for me. Visibility % is the better metric for AEO than any kind of “rank.”
Think about it this way. In traditional SEO, you could argue that “ranking #3” for a keyword told you something meaningful. But in AI search, there are no fixed positions. There’s no SERP with 10 blue links. When someone asks ChatGPT a question, the model generates a fresh response every time. Your brand might get mentioned in 7 out of 10 responses today, 3 out of 10 tomorrow, and 9 out of 10 next week.
A “rank” doesn’t capture that reality. Visibility percentage does. It tells you what share of AI-generated responses include your brand for a given topic. That’s the metric that matters and it’s the one agencies should be building their AEO reports around.
Now, this doesn’t mean that everything we learned in SEO gets thrown out. The revenue KPIs stay exactly the same: traffic, conversions, new customers, revenue. What changes are the leading indicators you track to predict and influence those outcomes. Visibility % across LLMs is the new leading indicator. Rankings on a traditional SERP are becoming a trailing one.
Track the Right LLMs for Your Clients
One of the mistakes I see agencies making is treating “AI search” as a monolith. It’s not. Each LLM has different training data, different grounding sources, and different biases. Your client’s brand might dominate on Perplexity but be completely invisible on Gemini. A blended “AI visibility” number without a per-model breakdown is hiding more than it’s revealing.
For most clients, you want to track ChatGPT, Gemini, and Perplexity.
B2B, enterprise, and companies selling to technical users or developers should always include Claude in their tracking.
Here’s what agencies need to know about each LLM:
- ChatGPT: This is the one every client asks about. It’s the most consumer-facing LLM and drives the most referral traffic for most verticals. If you’re only tracking one model, this is it. But you shouldn’t be tracking only one.
- Gemini: Google’s model is increasingly integrated into the Google ecosystem. For clients who depend on Google traffic (which is most of them), Gemini visibility matters because it influences AI Overviews and the direction Google search is heading.
- Perplexity: This is the one that’s gaining serious traction as a search replacement. Perplexity cites its sources, which means brands that show up here often see direct referral traffic. It’s especially important for informational and commercial-intent queries.
- Claude: If you’re working with B2B or enterprise clients, Claude is one to watch closely. Enterprise buyers and technical decision-makers are increasingly using Claude for research, vendor evaluation, and procurement decisions. We’ve seen this firsthand with our B2B clients at EWR Digital, Claude visibility correlates strongly with high-value lead quality in enterprise verticals.
- DeepSeek is another one that’s gaining ground in certain B2B and research-related circles. It’s not going to be relevant for every client, but for technology companies, SaaS brands, and enterprise-focused businesses, it’s worth monitoring.
The point is that you, as an agency, need to know your client’s buyer persona, understand which LLMs those buyers are using, and track accordingly. Don’t waste budget tracking models that don’t matter for a specific client, and don’t miss the ones that do.
Free Methods for Tracking & Reporting LLM Visibility
There are free ways to do this. I’ve used them. You probably will too, at least at first. Here are the most common approaches agencies start with:
- Manual spreadsheets: Open ChatGPT, Perplexity, Gemini, and Claude. Run a set of prompts. Check if the brand was mentioned. Log the results. Repeat monthly. It works when you have one or two clients. It doesn’t survive five.
- Looker Studio dashboards: You can build custom dashboards that pull in GA4 data to track LLM referral traffic. This gives you the traffic and conversion side of the equation, but it doesn’t tell you about visibility across the models themselves.
- Custom reports in GA4: Google Analytics 4 is getting better at identifying AI referral traffic. You can set up custom channel groupings and explorations to isolate traffic from chatgpt.com, perplexity.ai, and other AI sources. Again, useful for the traffic piece but it doesn’t measure upstream visibility.
- Custom scripts and APIs: If you’ve got a technical team, you can build your own tracking system using LLM APIs. Run prompts programmatically, parse responses for brand mentions, log results over time. I know agencies doing this. It works, but it’s a maintenance headache and it breaks every time a model updates its API.
These work to varying degrees for a small book of business. But if you’re truly trying to scale an agency and if you’re managing 5, 10, 50 clients then you need a tool that’s consistent, reliable, and fits into your existing workflows. The manual approaches eat margins, produce inconsistent data, and don’t scale.
Best AEO Tracking Tool for Agencies
The AEO tool landscape has matured a lot in the past year. There are now several legitimate options for agencies that need to report on LLM visibility. We actually published a full comparison of the best AEO tracking software tools on the EWR blog if you want the deep dive on each platform.
At a high level, the market includes SEO/AEO/GEO hybrid platforms like Semrush’s AI Toolkit (convenient if you’re already in their ecosystem, but pricey), Gumshoe AI (solid for usage-based reporting), Profound (strong for e-commerce), and a few others targeting enterprise use cases (see our review of Ahrefs Brand Radar as an example).
The tool I prefer to use for EWR clients and the one I strongly recommend to agencies is Cairrot.
I’ll tell you exactly why, because the reasons are specific and they matter for agency operations.
Why Cairrot Is My Preferred AEO Tracking Tool
I’ve tried several AEO tracking platforms on the market. I keep coming back to Cairrot because it was built by people who understand all the little features an agency might need to make beautiful AEO reports without making life harder than it has to be. Here’s what I mean:
- White-label reports and dashboards: This is non-negotiable for agencies. If you’re sending a client a report with another company’s branding on it, you’re undermining your own authority. Cairrot’s white-label capabilities let us deliver polished, EWR-branded reports that look like we built the entire system ourselves.
- Multi-client interface: One dashboard, all clients. I can switch between accounts without logging in and out. When you’re managing multiple brands, this saves hours every month. It sounds like a small thing until you’ve spent 20 minutes logging in and out of separate tool instances for each client.
- Wholesale pricing that scales: This is a big one. A lot of AEO tools use per-client pricing models that destroy agency margins as you grow. Cairrot’s wholesale pricing means our cost per client goes down as we add more accounts. That’s the kind of pricing structure that actually supports agency growth instead of punishing it.
- API and MCP integrations: Some of our clients want data piped into their own BI tools or internal dashboards. Cairrot’s API makes that possible. The MCP (Model Context Protocol) integration is newer, but it’s positioning Cairrot for the agentic workflow future that’s coming. Agencies that automate early are going to win.
- Custom reporting dashboards: Not every client needs the same report. Cairrot lets us customize what each client sees, which metrics are front and center, and how the data is presented. That flexibility is essential when you’re serving clients across different verticals with different priorities.
- GA4 and Google Search Console integration: AEO data in isolation is interesting. AEO data correlated with actual search behavior and traffic is actionable. This integration bridges the gap between “are we visible in AI?” and “is that visibility driving business results?”
My Favorite Report Types in Cairrot
Let me get specific about the reports I actually use with clients, because I think this is where a lot of agencies get stuck. They know they need to report on AEO, but they don’t know what the report should look like.
Here are the four report types I lean on most heavily:
Heatmap Reports: These are probably my single favorite visualization in Cairrot. The heatmap shows visibility intensity across prompt topics over time. Hot spots mean strong, consistent mentions. Cold spots flag areas where the brand is losing ground or never had visibility. When a client asks “what should we focus on next?” I point at the cold spots. It maps directly to content strategy and makes the next steps obvious.
Leaderboard Reports: For any given prompt topic, how does my client’s brand stack up against competitors? The leaderboard gives you share of voice per topic, per LLM. It’s the competitive positioning slide in every client deck. When a client asks “how do we compare to [competitor]?” This is the answer. It’s concrete, visual, and impossible to argue with.
Sentiment Comparisons Against Competitors: This goes beyond just “are we mentioned?” to “what is the AI saying about us versus the competition?” Sentiment analysis across LLM responses tells you whether your brand is being positioned favorably or unfavorably relative to competitors. That insight drives messaging strategy, content priorities, and even product positioning conversations.
AI/LLM Traffic Dashboards via GA4 Integration: This is the bridge between visibility metrics and business outcomes. Through Cairrot’s GA4 integration, we can show clients not just where they’re visible in AI responses, but how much traffic and how many conversions that visibility is actually driving. This is the report that turns AEO from a “nice to know” into a revenue conversation.
These four report types together tell a complete story: here’s where you stand (heatmap), here’s how you compare (leaderboard), here’s what AI is saying about you (sentiment), and here’s what it means for your business (GA4 traffic). That’s an AEO report a client can actually use to make decisions.
How Often Should Agencies Report on AEO?
Based on what’s working for us at EWR Digital and what I’m hearing from other agencies on The Best SEO Podcast, here’s the cadence that makes sense:
Monthly: Send an automated visibility report. Keep it focused. Top-level metrics, heatmap snapshot, a brief narrative of what moved and why. This keeps clients informed without eating your team’s hours. Most of this can be automated or templated in Cairrot.
Quarterly: Deliver a deep-dive analysis. This is where you pull in leaderboard reports, sentiment comparisons, GA4 correlation data, and strategic recommendations. It’s the “here’s what’s working, here’s what’s not, and here’s what we’re doing next” conversation.
Why this cadence works? AI search results for high-intent, bottom-funnel queries, the ones that actually drive revenue and don’t fluctuate as wildly as broad informational queries. Monthly is enough to catch meaningful movement. Quarterly is where you build and adjust strategy.
One trap to avoid is over-reporting on top-funnel informational queries. These bounce around constantly in LLMs, and reporting on every swing creates noise that distracts from what matters. Focus your reporting on the prompts and topics that correlate closest with revenue. In SEO language, that means commercial and transactional intent queries.
Start AEO Now and Get Visible in LLMs Before Your Competitors Do
The agencies that build AEO reporting into their stack today are going to own the conversation with clients before their competitors figure out what’s happening. I’ve seen this play out before and it’s the same dynamic that happened with SEO reporting 15 years ago, and with content marketing reporting a decade ago. The early movers set the standard. Everyone else plays catch-up.
The shift is clear: from rankings to visibility. From individual keywords to prompt topics. From single-platform reporting to multi-LLM analytics. The underlying business KPIs haven’t changed. The measurement layer has.
If you want to learn more about how we’re approaching AEO tracking and LLM Visibility at EWR Digital, or if you want help getting started with AEO reporting for your agency or your brand: reach out to our team. We’re helping agencies and enterprise brands build the reporting infrastructure they need to compete in AI search, and we’d love to show you what’s possible.