AI has changed how brands get discovered, but most teams are still measuring the wrong things. Here’s a practical framework for the tools and metrics that actually matter in the era of generative search.
Updated September 2026: This guide was first published in March 2026. It has been revised to reflect how quickly the market has matured, such as the growing recognition that being mentioned by AI and being recommended by AI are different outcomes — along with updated tool details. Disclosure: Aperture Insights now offers its own AI visibility product, aimprint™, which is included in the comparison below.
In our first two installments, we explored the fundamental shift AI has brought to media strategy and established the Trust Architecture required to win in an algorithmic world. But for the modern marketing and communications professionals, understanding these shifts is only half the battle. A question I am frequently asked by marketing and communications leaders is: “How do I actually prove we are winning?”
With traditional search, we had the luxury of clear metrics: click-through rates, keyword rankings, and domain authority. In the era of the Large Language Model (LLM), the “black box” has become more opaque. When a user asks Perplexity or ChatGPT for a recommendation, there is no “page two” of results.
When this post first ran, we put it this way: you are either cited, or you are invisible. Six months later, that framing looks incomplete. Appearing in an AI answer is only the first hurdle. A brand can show up in nearly every relevant response and still lose, because a competitor is the one the answer actually recommends.
To manage what we measure, we need a new toolkit. This post provides a framework for the emerging landscape of AI visibility tools and a guide to help you choose the right partner for your measurement journey.
From Being Found to Being Chosen
Measurement in AI is fundamentally different because the “search engine” is no longer just a pointer to a website; it is an answer engine. Many AI tools leverage retrieval-augmented generation (RAG) to pull real-time external sources into their responses. Optimization is no longer just about being found; it’s about being trusted as a primary source.
We are seeing a shift from SEO (Search Engine Optimization) to GEO (Generative Engine Optimization): the optimization of content, structured data, and brand signals to increase visibility and citation frequency in AI-generated search results. In this new paradigm, success is best measured as a progression::
- Presence: Does your brand appear at all? Mention rate captures the share of relevant prompts where your brand is named.
- Share of voice: When you appear, how much of the conversation do you own relative to competitors? Answer Share of Voice (ASOV) measures your share of brand mentions across a set of prompts.
- Citation share: Is your brand cited as a source — content AI trusts enough to reference?
- Recommendation: When AI compares options, does it actually recommend you, or mention you on the way to recommending someone else? This is the outcome that drives decisions, and the one most tools don’t measure.
- Narrative: Is AI describing your brand accurately, using the key messages you’ve spent years refining?
| Traditional SEO | Generative Engine Optimization (GEO) | |
|---|---|---|
| Core Philosophy | Being Found | Being Trusted |
| Primary Goal | Ranking on “page one” of results | Earning the recommendation, not just a mention |
| Success Metrics | CTR, keyword rankings, and domain authority | Mention rate, share of voice, citation share, recommendation rate, and narrative accuracy |
| User Experience | Browsing a list of potential links | Receiving a synthesized, direct recommendation |
| The “Black Box” | Transparent (mostly): clear metrics and rankings | Opaque: “Answer share” across varying LLM responses |
In short, traditional SEO is about being found in a list of links; GEO is about being trusted enough to be cited in a direct answer.
A Framework for AI Visibility Tools
The market for AI tracking is maturing rapidly. To make sense of it, it helps to think in terms of what the tools do, not what their marketing pages call them. We can generally group these tools into the following archetypes:
1. AI answer trackers (LLM “SERP monitors”)
These tools run panels of prompts across AI engines, then show you how often your brand appears, in what position and context, and which URLs and domains are being cited.
They answer questions like “What does ChatGPT say when someone asks about top providers in our category?” or “How often do we show up in Google’s AI Overviews?” The more sophisticated trackers now go further, distinguishing a mention from a recommendation and grouping prompts by buyer intent.
Examples: Am I On AI?, Keyword.com
2. AI narrative intelligence platforms
These tools don’t just track LLM answers; they blend that data with news, social, and other media signals to show how narratives travel across channels.
They answer questions like “Is the story AI tells about us aligned with what journalists and creators are saying?” and “Are early shifts in AI answers a leading indicator of a broader reputation issue?”
Example: Meltwater GenAI Lens
3. AI optimization and experimentation tools
Monitoring tells you what’s happening; optimization tools help you change it.
In March, we called this the least mature category. It has moved quickly since, and now splits into two distinct approaches:
- Diagnostic optimization identifies where you’re losing visibility and categorizes the likely remedy — for example, flagging that a gap stems from missing editorial coverage (a PR problem) rather than your own website (a content problem). The work of closing the gap stays with your team.
- Generative optimization goes further, drafting the content itself — articles, FAQ sections, outreach emails — and in some cases pushing it to your CMS for approval.
The distinction matters. Diagnostic tools keep your team in control of what gets published under your brand; generative tools trade some of that control for speed.
Examples: Peec AI’s Actions (diagnostic); Profound’s Agents (generative)
How Popular Tools Line Up
With that framework in mind, here’s how some tools map to those roles. This is not an exhaustive list, nor is this intended as a product review. Rather, this list represents a cross‑section of popular platforms that illustrates how the landscape breaks down.
Am I on AI?
Am I On AI? is a focused tool that shows how businesses rank on AI platforms with brand monitoring, competitor ranking, prompt tracking, and source analysis.
Because it is lighter‑weight and more accessible than many enterprise platforms, it’s a helpful starting point for smaller teams who need to monitor AI visibility but can’t yet justify a full GEO stack. It is a clear AI answer tracker designed for scrappier use cases.
Am I On AI is currently limited to ChatGPT with weekly scanning frequency.
Peec AI
Peec AI focuses on measuring AI search visibility across models with an emphasis on competitive benchmarking. It tracks how often your brand appears in AI-generated answers, organizes prompts by topic and intent, and identifies which sources are driving citations for you and your competitors.
Its Actions feature, included on all plans, clusters visibility gaps by source type — editorial, community, reference sites, your own website — and scores each as an opportunity, turning raw tracking data into a prioritized list of what to address. It’s a data-rich tracker best suited to teams that want granular competitive intelligence at an accessible price point.
Plans start at 50 prompts across three AI engines of your choice, with additional engines available as paid add-ons — worth factoring in if you need full engine coverage, since the headline price covers three.
Ahrefs Brand Radar
Ahrefs recently launched Brand Radar to tackle the “black box” of LLMs. Its standout feature is its massive database of real-world user prompts (200M+ total monthly prompts). Instead of just tracking what you think people are asking, Brand Radar shows you where your brand appears across millions of actual conversational queries. It is a premium, data-heavy tool for those who want to see citations across ChatGPT, Perplexity, and Gemini in one view. It also includes visibility in Reddit, YouTube and TikTok.
Pricing starts at $199/month per platform or $699 for all available platforms: Google AI Mode/AI Overviews, ChatGPT, Perplexity, Copilot, Gemini.
Revere
Revere approaches AI visibility from the brand management side. What sets it apart is its proprietary Revere Brand Index Score (RBI), which synthesizes the likelihood that an LLM will recommend your brand and whether that representation will be favorable. The RBI combines multi-prompting, attribute analysis, and sentiment scoring to give marketers a single metric for tracking brand affinity over time.
Beyond monitoring, Revere offers AI brand audits and what it calls “affinity optimization” — identifying content and actions to improve how LLMs portray your brand. It’s best suited for brands focused on qualitative AI perception — not just whether you appear, but how you’re described.
Meltwater GenAI Lens
For PR professionals already using Meltwater, their GenAI Lens is the most logical step. It bridges the gap between traditional media monitoring and AI search, allowing you to see how a piece of earned media in The New York Times eventually feeds into an LLM’s knowledge base.
It monitors brand, product, and competitor mentions across most major LLMs and assistants, including ChatGPT, Gemini, Perplexity, Claude, Grok and Deepseek. It shows both how your brand is represented —including sentiment analysis, key phrases and other brands/people mentioned— along with where the models are sourcing their information.
By pairing AI answers with news and social data, it helps comms teams catch reputational risks earlier and see whether AI is amplifying or muting key narratives. This is a textbook AI narrative intelligence solution that slots into familiar PR workflows.
GenAI Lens is typically an add-on to other Meltwater solutions. As such, pricing is custom.
aimprint™ (Aperture Insights)
aimprint measures brand presence in AI answers across three dimensions: Reach (how often your brand appears across relevant prompts), Influence (whether it’s recommended or merely mentioned), and Narrative (whether AI characterizes it accurately relative to your intended positioning). Seven primary KPIs sit beneath those pillars — Mention Rate, Answer Share of Voice, and Citation Share; Recommendation Rate and Competitive Win Rate; Answer Sentiment and Narrative Alignment — rolling up into a composite 0–100 aimprint Score. Results are tracked across up to six AI engines depending on plan, mapped to the prompt level, and filterable by engine and by branded versus unbranded prompts.
Its distinguishing emphasis is the depth behind the recommendation question. Rather than stopping at whether a brand is mentioned or considered, it measures whether AI actually picks it — including which brand wins when several appear in the same answer.
aimprint offers two engagement models. Self-directed plans provide the interactive dashboard, with pricing published openly rather than behind a sales call. The Managed plan adds a practitioner layer: a senior analyst reviews the data each month and delivers a written report with strategic recommendations — the “so what” and “now what” alongside the numbers.
For agencies, the same structure doubles as a client-reporting tool. A single user can work across multiple client brands, and the KPIs map to what marketing and PR teams already report on — reach, share of voice, key message pull-through. The composite score gives a consistent number to track across an engagement. It measures progress over time; attributing that movement to a specific campaign still takes judgment, since AI answers also shift with model updates and competitor activity.
Disclosure: aimprint is an Aperture Insights product.
Profound
Profound is the most prominent enterprise-level platform in the category and among the best funded. It provides granular data on how your brand is mentioned across major AI search tools.
It maps the “information diet” of AI engines — which sources they cite, where you lose out to competitors, and how sentiment and framing change over time. Its Agents go well beyond measurement: they can draft articles, FAQ content, and outreach emails, build PR target lists from the domains AI actually cites, and push drafts to your CMS behind an approval step.
The entry tier includes 50 prompts tracked daily on ChatGPT only, with limited functionality. The mid-tier adds three engines and 100 prompts with most core functionality. Full functionality, including broader engine coverage, requires custom enterprise pricing.
Beyond the Dashboard: The Human as Interpreter
AI visibility measurement is often framed as a purely technical problem. It isn’t.
Even the best platforms cannot fully interpret subtle brand positioning, narrative framing and competitive context. A brand might appear frequently in AI responses but in the wrong context. Citations may come from low-trust sources, or competitors may control the narrative framing. These judgments still require human interpretation.
In practice, the most effective teams combine:
- Automated measurement
- Strategic analysis
- Editorial judgment
Most AI visibility tools reduce the manual work; they answer the “What?”. Even the newer optimization features that suggest remedies don’t replace strategic thinking — the “So what?” and “Now what?”. As platforms become more sophisticated and integrated into workflows, the role of comms analysts is shifting from Reporter to that of Strategic Interpreter.
Where to start
If you’re on a small team or tight budget, start with Am I On AI? or Keyword.com. Both offer affordable entry points and will give you a baseline understanding of your brand’s AI presence. Before committing to any platform, take advantage of free trials where they’re offered — but treat them as a way to evaluate fit rather than a true preview, since trials often limit prompts, engines, or features compared with the paid product. Use what you learn to build a business case for deeper investment.
If you’re an agency managing multiple brands, Peec AI’s credit-based model and Keyword.com’s unlimited projects offer flexible scaling. Profound’s enterprise tier is worth evaluating for high-value accounts where the ROI justifies the price.
If you’re already using Meltwater or Semrush, activate their AI visibility modules first. The integration advantage is real—you’ll get AI data in context alongside the metrics you already track, with no new vendor relationship to manage.
If you want to know whether AI recommends you, not just mentions you, aimprint™ is Aperture Insights’ own product, built directly in response to the gaps described in this post. It produces a composite aimprint Score across Reach, Influence, and Narrative, with competitive benchmarking, visibility gap analysis, and transparent pricing. Teams that want interpretation as well as data can choose the Managed plan, which adds a monthly analyst review with strategic recommendations.(Disclosure: aimprint is our product.)
| Platform | Best For | Key Strength | Delivery Model |
|---|---|---|---|
| AI Answer Trackers | |||
| Am I On AI? | Teams just starting their AI visibility journey | Fast presence checks (ChatGPT-focused) | Self-serve SaaS |
| Peec AI | Data-focused teams wanting granular competitive monitoring | Competitive benchmarking plus gap prioritization via Actions (6+ LLMs) | Self-serve SaaS |
| Ahrefs Brand Radar | Data-driven teams who want to see citations across a massive real-world prompt database | Database of real-world user prompts (6+ LLMs) | Self-serve SaaS |
| aimprint™ (Aperture Insights) | Teams that want to know whether AI recommends them, not just mentions them | Recommendation-level measurement with a composite aimprint Score (up to 6 LLMs) | Self-serve SaaS; Managed plan adds a dedicated analyst |
| AI Narrative Intelligence | |||
| Revere | Brands focused on qualitative AI perception and attribute analysis | Narrative intelligence (5 LLMs) | Managed / custom |
| Meltwater GenAI Lens | Agency and in-house teams who want AI integrated with PR tracking | Integrated media + AI monitoring (8 LLMs) | Managed / add-on to existing contract |
| AI Optimization | |||
| Profound | Enterprise teams wanting end-to-end AI optimization and automation | Deep tracking plus content-generating Agents (10 LLMs) | Self-serve SaaS (enterprise support available) |
What’s Next: Ecosystem and orchestration layers
Beyond the tools themselves, there are standards focused on integration, which will allow users to focus on interpretation rather than the tedious manual work of stitching together screenshots and spreadsheets.
The Model Context Protocol (MCP) is an emerging open standard that allows AI models to connect directly to external tools and data sources. For comms teams, the practical implication is that an AI assistant could eventually pull from your visibility data, media monitoring, CRM, and analytics in a single workflow—reducing the manual work of stitching together insights from separate dashboards.
The practical benefit for comms leaders is simple: less time gathering and organizing data and more time understanding what’s happening and acting on it.
Practical Next Steps for Comms and Marketing Teams
Whether you’re in‑house or at an agency, a practical sequence looks like this:
1. Define the questions before the tools
Decide what matters most right now: baseline visibility, competitive recommendation, reputation risk, or campaign impact. That will narrow the category of tools you actually need.
2. Start with a baseline panel
Identify 20–50 queries that reflect your customers’ real questions, across key markets and personas. Use an AI answer tracker to establish your current visibility and narrative — and, where the tool supports it, separate the prompts where you’re merely mentioned from those where you’re recommended.
3. Connect AI answers to existing signals
If you already have Meltwater or similar tools, explore how GenAI Lens or equivalent features can pull AI into the same place you look at news and social. Align AI visibility metrics with the KPIs you already track for brand health, share of voice, or crisis detection.
4. Diagnose before you intervene
Before choosing a tactic, identify why you’re missing. A gap rooted in editorial coverage calls for PR outreach; one rooted in your own site calls for content or structured-data work; one rooted in community discussion calls for authentic engagement. Then test specific interventions — new explainers, FAQ pages, structured data updates, proactive PR placements — and measure whether they shift not just presence, but recommendation.
5. Plan for orchestration, not just another dashboard
As MCP and similar approaches mature, ask every vendor how they plan to integrate AI visibility data into your broader analytics stack. The goal is a living, cross‑channel understanding of brand narrative health—not yet another tab to check in the morning.
AI has already changed how people discover and evaluate brands. The advantage goes to teams that treat AI visibility as a measurable channel, build the right mix of tools around it, and keep human judgment firmly in the loop.
Final Thoughts
Measurement in the AI era is not a “set it and forget it” task. The models update their weights, the search engines change their citation styles, and new protocols like MCP emerge. By building measurement capabilities now, you aren’t just tracking your current performance—you are laying the foundation to survive the most significant shift in media history.
SEO remains the baseline for being found. The brands that get ahead are the ones that adapt early, pairing strong search fundamentals with GEO strategies built for visibility, trust, citation and recommendation..
Frequently Asked Questions
SEO (Search Engine Optimisation) focuses on ranking in traditional search results — the goal is clicks and traffic. AEO (Answer Engine Optimisation) targets AI-powered features like Google AI Overviews and voice search, where your content is surfaced as a direct answer — no click required. GEO (Generative Engine Optimisation) goes further: it’s about being cited as a trusted source when users query AI tools like ChatGPT or Perplexity. Think of it this way — SEO gets you on the list, AEO gets you selected as the answer, GEO gets you cited as the source.
Being mentioned means your brand is named somewhere in an AI answer. Being cited means your content is used as a source — the page AI references to build its answer, often shown as a link. Being recommended means the answer actually points the user toward you as the best fit or the brand to choose.
These don’t always move together. A brand can be named in nearly every relevant answer while a competitor consistently gets the recommendation. A brand can even be cited as a source while the answer built from its content recommends someone else — in effect, doing the AI’s research for a competitor. For most brands, recommendation is the outcome that drives decisions, which is why measuring mentions or citations alone can create a false sense of security.
For most brands, a weekly baseline check is a good rhythm — enough to spot trends without creating noise. That said, treat it like a post-campaign check-in: if you’ve published major content, earned significant press coverage, or launched a new product, pull your visibility data shortly after to see if it moved the needle. AI models update their training data on different schedules, so consistency matters more than frequency.
These two metrics measure different dimensions of AI visibility. Answer Share of Voice (ASOV) measures breadth and presence — it captures the percentage of relevant prompts where your brand is mentioned at all, whether cited or simply named. Citation Share measures depth and authority — it tracks how often your brand is cited as a source, signaling that AI models trust your content enough to reference it. Together, they give you the full picture: ASOV tells you whether AI knows you exist; Citation Share tells you whether AI trusts you.
Retrieval-augmented generation (RAG) is the mechanism by which AI systems like Perplexity and ChatGPT pull real-time external sources into their responses. Rather than relying solely on their training data, these systems retrieve and cite live web content — which is why being recognized as a trustworthy, citable source is central to any GEO strategy.
No. The manual approach — running targeted queries in ChatGPT, Perplexity, or Gemini and tracking responses in a spreadsheet — costs nothing and is a perfectly valid starting point. Purpose-built GEO platforms (such as Semrush’s AI toolkit, or Profound) add automation and benchmarking, but pricing varies considerably. Start manual to learn what questions matter for your brand, then invest in tooling once you know what you’re measuring.