Lumar GEO and AEO Toolkit Review: Checking Brand Mentions in AI Answers

What is in Lumar's GEO toolkit?

Lumar groups its GEO features into three stages, as described on its product pages:

• AI discovery: can AI systems reach and read your pages? Reports cover availability, crawlability, raw versus rendered HTML, robots.txt issues, nosnippet directives and AI bot activity on your site.
• AI understanding: can they make sense of the content? Reports cover page content and structured data.
• AI inclusion: are your pages relevant and trusted enough to be used? Reports cover semantic relevance and AI referral traffic, with freshness and authority analysis under development.

GEO health scores and dashboards summarise issues for stakeholders, and Lumar can draft developer tickets for each fix. The same platform covers technical SEO, site speed and accessibility, with monitoring, alerts and pre-release testing that plugs into deployment pipelines.

How does Lumar check brand mentions in AI answers?

Lumar describes four brand-level checks:

• AI brand presence, measured with blind query tests. Models are asked about a topic without your brand in the question, showing whether they name you on their own.
• AI brand sentiment, analysing how models describe you on trust, authority and quality, with a breakdown by model.
• Google AI features tracking, showing whether you appear in AI Overviews and AI Mode.
• Content uniqueness, comparing your pages with top-ranking competitors to show what only you say.

Blind queries are the right method for measuring unprompted awareness, which is what matters when buyers ask for recommendations. Branded questions, such as "what is [brand]?", test how you are described, not whether you are chosen. A fuller method for this kind of counting is in how to count brand mentions in AI responses.

Since March 2026 Lumar has also sold AI Search Visibility, a prompt tracker inside the same platform. You set topics and your own prompts; Lumar runs them on ChatGPT, Google AI Overviews, AI Mode, Gemini, Perplexity and Claude at a cadence you choose, and reports an AI Visibility Score, brand sentiment, citation positions and competitor share of voice, with up to two years of history. Prompt allowances are not published, so ask how many prompts your quote includes.

What does the GEO Content Evaluator do?

The Content Evaluator scores pages with what Lumar calls LLM-as-a-judge methods, the same approach researchers use to have one AI model grade text. Its parts:

• Query to question mapping: it takes real Search Console queries and maps them to likely AI prompts, then scores how well your content answers them.
• Content precision and semantic relevance scores, showing where pages cover a topic well and where they miss what users want.
• Authority and trust audits, highlighting expertise and trust signals.
• Action plans, prioritised suggestions for each page.

Grounding prompts in Search Console data is a sensible way to avoid tracking questions nobody asks. The caution is that the scores are an AI's judgement of your page, not a measure of whether AI engines cite it. Lumar does not publish evidence that higher scores lead to more citations, so test the link on your own pages before relying on the scores.

What is Lumar best at?

Connecting visibility gaps to causes. When a page is not cited, the reason is often technical: content that only appears after scripts run, a robots rule blocking an AI crawler, a nosnippet tag, or duplicate pages splitting signals. Lumar crawls at high speed, up to 450 URLs a second in its lab tests, and handles very large sites, so it can find these problems across millions of pages and turn them into tickets. Mention trackers rarely do this. The rendering issue in particular is covered in how indexing, canonical and schema issues prevent LLM citations.

Where does it fall short?

• Prompt allowances. Prompt counts per package are not published.
• No public pricing. Lumar sells by quote; it is an enterprise purchase.
• Monitoring depth. Lumar's tracker covers competitor share of voice and citation positions, but it is newer than dedicated trackers; ask to see raw answers and cited URLs for a few of your prompts before relying on its scores.
• Proprietary scores. Health and content scores are Lumar's own measures and need validating against real citations.

We have not tested Lumar ourselves; this review is based on its published pages.

Who is it for?

• Good fit: enterprise SEO teams with large, complex or script-heavy sites, especially those already using Lumar for technical SEO.
• Reasonable fit: teams that pair a dedicated mention tracker with a technical platform, and want both views of the same pages.
• Poor fit: small sites, and teams whose main question is which competitors AI recommends and why.

Mention-first trackers are covered in the Rankscale review and the Peec AI review.

How can you run a blind brand check yourself?

Before buying any tool, you can test unprompted awareness by hand:

1. Write 20 topic questions buyers ask, without your brand or competitors in them.
2. Ask each in ChatGPT, Gemini, Perplexity and Google's AI Mode, in a fresh session.
3. Run each question three times, since answers vary.
4. Record whether you are named, in what position, and which sources are cited.
5. Compare with two competitors on the same questions.

If you are rarely named, check the technical basics next: can AI crawlers fetch your key pages, and does the content appear without scripts? That is where a crawler such as Lumar earns its place. Finding the right questions is covered in how to find prompts that mention your brand.

What might this look like in practice?

Consider an illustrative travel insurance company, Wayfarer Cover, with 80,000 pages. Blind checks showed it rarely named for questions about cover for pre-existing conditions, despite detailed guides. A Lumar crawl found those guides built from script-loaded components, so the raw HTML that many AI crawlers read held only headings. The team moved the guide text into the server-rendered page and fixed a robots rule that blocked one AI crawler from the guides folder. Three months later, the guides were cited in several assistants' answers.

What do people ask about Lumar's GEO toolkit?

Does Lumar track brand mentions in ChatGPT?

Yes. AI Search Visibility tracks your prompts in ChatGPT, Gemini, Perplexity, Claude and Google's AI Overviews and AI Mode.

Is Lumar the same as Deepcrawl?

Yes. Deepcrawl became Lumar.

How much does Lumar cost?

Lumar prices by quote, with packages built around the features you need.

Can Lumar see AI bots on my site?

Yes. Its AI discovery reports include AI bot interaction with your site.