Track how competitors appear alongside your brand in AI-generated answers and benchmark your mention share.

Overview

AEO Optima's Competitor Intelligence features let you track how competitor brands are mentioned in the same AI responses where your brand appears — or should appear. Instead of monitoring your brand in isolation, you get a complete picture of the competitive landscape within AI answer engines.

Understanding who AI models recommend alongside (or instead of) your brand is essential for building an effective AEO strategy.

Key Metrics

Three core metrics power the Competitor Intelligence view.

Mention share (previously called share of voice) measures your brand's mentions as a percentage of all brand mentions detected across your snapshots. It answers the question: "When AI models talk about brands in my space, how much of that conversation includes mine?"

For example, if your snapshots contain 100 total brand mentions across all brands, and your brand accounts for 35 of those mentions, your mention share is 35%.

What counts in the denominator: your mentions ÷ (your mentions + every tracked competitor's mentions + the most-mentioned emerging brands we detect in the same answers). Emerging brands are names that surface repeatedly in AI answers even though you haven't added them as competitors yet — including them keeps mention share honest, so a new entrant taking share shows up as a decline in yours rather than being invisible. The denominator is the total brand conversation, not the total number of snapshots — a response that mentions no brands at all doesn't dilute your share.

How emerging brands are detected (LLM-as-judge NER): an LLM reads each AI response and lists only the proper-noun organization names actually mentioned — companies, brands, products, agencies, services. The judge is given explicit negative examples drawn from real noise we observed: common English words ("the", "paid", "marketing"), places ("Middle East", "Abu Dhabi"), section headings ("Overview", "Key Features"), generic categories ("Agencies", "Solutions"), and measurements ("7 km of Blue Flag beaches") are all excluded. The judge runs at temperature 0 so the same input always produces the same brand list. If the judge is briefly unavailable, the older regex-based extractor stands in as a safety net so capture never fails — these rows are tagged differently in our database so we never confuse the two.

A company's mentions are the answers that named it. Two things read every AI answer: the matcher that runs at capture time, and the LLM judge that lists the organisations an answer mentioned. Either one seeing your competitor is enough — the answer counts once, for that company. We do not add the two tallies together (that would count the same answer twice) and we do not take the larger (that would throw away the answers only the other one saw). On a live account this is the difference between reporting 76 mentions for a rival and reporting the 226 answers that actually named them.

Curating an alias moves that company's count everywhere, back through its whole history. Names are resolved to companies each time a chart is drawn, using the aliases as they stand today. So the moment you tell us that "JSW Neosteel" is the company you added as "JSW Steel", every answer that ever named the product line joins that company's count — on the chart, in the per-model breakdown, in the trend and in the export, for the whole period you are looking at, not just from today onward. Last month's chart changes to today's understanding of who is who. That is deliberate: a chart should show what you now know, not what we guessed before you told us.

A company counts once, whatever it is called. AI answers spell the same company several ways — "Tata Tiscon" and "Tatatiscon", or a product line like "JSW Neosteel" for the company you added as "JSW Steel". Before a name is treated as an emerging brand we check whether it is a company you already track: first by matching the name itself, then against the aliases you have curated on the Competitors page. If it is one you track, its mentions stay with that competitor and no second row appears. Only genuinely new companies are added as emerging brands. This is also why curating an alias is worth doing — it is how you tell us that a product line or a parent company is the rival you already listed.

A company you don't track has to be named twice before we list it. Companies you have added as competitors are always shown, however rarely they appear — you told us they matter. For everyone else there is a bar: the company has to have been named in at least 2 answers in the window, and we show the 20 most-named of those. This is not a small trim. On a live account, 642 different company names appeared across 544 answers and 386 of them — sixty per cent — were named in exactly one answer. Listing those as rivals would bury the handful that actually compete with you. Both the mention-share chart and the Competitors page apply the same bar, so the two always agree, and the caption on each says how many companies cleared it.

Competitor names shown are the ones you curated. Where you have given a competitor a proper name, that name is what appears — on the chart, in the per-model breakdown, in the trend legend and in the CSV export, all together. The underlying data still keys on the name captured at the time, so history stays intact.

A rising mention share means your brand is gaining prominence relative to competitors. A declining mention share means competitors are being mentioned more frequently — even if your absolute visibility stays the same.

Competitor Mentions

The raw count of how many times each competitor brand was mentioned across all snapshots in the selected date range. This metric helps you understand:

Which competitors AI models mention most frequently.
Whether certain competitors are gaining or losing AI visibility.
Which brands the AI models consider your closest alternatives.

Brand Overlap (%)

Brand Overlap measures how often your brand and a specific competitor appear in the same AI response. A high overlap percentage means AI models frequently mention both brands together — often in comparison or "alternatives" contexts.

For example, if a competitor has a 70% overlap with your brand, it means that in 70% of the AI responses where that competitor appears, your brand also appears (and vice versa).

High overlap can be either positive (your brand is always in the conversation) or concerning (the AI model may be positioning the competitor as a direct substitute). The context depends on your sentiment and rank data.

Setting Up Competitor Tracking

Adding a Competitor

Navigate to Competitors from the sidebar.
Click Add Competitor.
Enter the competitor's brand name as it would appear in AI responses.
Optionally enter the competitor's website URL to help with identification.
Click Save.

Automatic Detection

Once a competitor is added, AEO Optima automatically detects mentions of that brand in:

All future snapshots — Every new snapshot is analyzed for competitor mentions as part of the standard analysis pipeline.
Existing snapshots — Previously captured snapshots are also scanned, so you get historical competitor data immediately without needing to re-capture.

No additional configuration is needed. The system handles detection across all AI models and all prompts.

Choosing Which Competitors to Track

Start with your most direct competitors — the brands that a potential customer would consider alongside yours. Common approaches include:

Direct competitors: Companies offering the same product or service category.
Aspirational competitors: Market leaders you want to be compared against.
Emerging competitors: Newer brands that are gaining traction in AI responses.

You can add or remove competitors at any time without affecting your existing data.

Which prompts mention this competitor — A list of your prompts where this competitor has appeared in AI responses. This reveals exactly which topics and questions are driving competitor visibility.
Which AI models mention this competitor most — A per-model breakdown that shows if a competitor is prominent across all models or concentrated in specific ones.
Mention trend over time — A mini chart showing whether this competitor's mentions are increasing or decreasing.

Tip: Click a competitor row to expand and see which prompts mention them. This reveals content opportunities — if a competitor shows up on a prompt where your brand does not, you know exactly which topic to address in your content strategy.

Strategic Use of Competitor Data

Identifying Content Gaps

The most actionable use of competitor data is finding prompts where competitors are mentioned but your brand is not. These are direct content opportunities:

Filter to a specific competitor.
Review the prompts where they appear.
For each prompt where your brand is absent, evaluate whether you should be mentioned.
Create or optimize content that addresses that topic, making it more likely that AI models will include your brand in future responses.

Monitoring Competitive Shifts

Track mention share over time. If a competitor's share is growing:

Check whether they've published new content recently.
Review the specific prompts driving their increased mentions.
Assess whether the shift is across all AI models or specific to one.

Benchmarking Progress

Use competitor data to set concrete goals:

"Increase mention share from 25% to 35% within 90 days."
"Appear in AI responses for at least 3 of the 5 prompts where Competitor X currently appears but we do not."
"Reduce Brand Overlap with Competitor Y below 50% by differentiating our positioning."