The way to Evaluate Your AI Visibility Towards Your Rivals

If a competitor is talked about extra typically, cited extra incessantly, or framed extra favorably in AI solutions, they win consideration and belief earlier than searchers even know you exist.

This information reveals you how you can run an AI search competitor evaluation that reveals:

• Which manufacturers in your {industry} dominate AI search right now
• The place your opponents earn visibility, however you don’t
• Which matters, pages, and sources drive their benefit
• What to prioritize to shut the hole quick

It’s a step-by-step course of you may repeat quarterly, or at any time when a brand new competitor, mannequin replace, or inner precedence change (akin to once you launch a brand new product) hits your radar.

1. Discover direct and significant opponents in AI search 

There are three primary sorts of opponents you compete with for AI visibility:

• Those who AI mentions you alongside
• Those who AI cites as a supply as a substitute of you
• Those who your viewers compares to you when utilizing AI search

You could find all three in Ahrefs’ Model Radar. As an example, when you enter your model, you should use the AI counsel characteristic to record different comparable manufacturers to yours:

These will usually be opponents you’re conscious of and that you simply compete with for model mentions in AI responses.

In case you take a look at the Search queries report, you’ll be capable to discover the opponents that your viewers incessantly compares you to. You’ll be able to filter the info by searching for queries that include your model and phrases with comparative intent, like “vs”:

For instance, Honda is a powerful competitor of Toyota’s that searchers incessantly evaluate:

Different manufacturers which might be generally in comparison with Toyota embody Nissan, Mazda, Ford, Lexus, Suzuki, and Subaru:

To search out opponents which might be cited as sources, take a look at the Cited domains report.

As a rule, you’ll see basic websites like Wikipedia, Reddit, or social media within the high spots. These websites are too giant and authoritative to check in opposition to in most instances.

However you’ll additionally spot different websites in your {industry} that get cited greater than your model. Generally they’ll be {industry} publications. Different occasions, they’ll be direct opponents, like Ford, Nissan, and Honda, within the instance beneath:

Choose those that take advantage of sense on your evaluation or that your stakeholders care in regards to the most.

2. Make clear the model entities you’ll evaluate for the evaluation 

After you finalize the record of manufacturers and web sites to check on your competitor evaluation, you’ll must outline the entities you’d prefer to obtain information for. Entities are the core folks, manufacturers, merchandise, and matters that serps can uniquely establish and hook up with what you are promoting.

You are able to do this within the search bar by clicking the choice so as to add variations as you enter every model:

For a easy evaluation, add the principle model and its area:

Do that on your model and every of your opponents.

For a extra complete evaluation, it’s also possible to add all merchandise, sub-brands, and different owned web sites to those fields. This provides you with deeper insights into every model’s visibility at an in depth degree.

For instance, for Honda, you might embody merchandise which might be looked for immediately, such because the Honda CR-V, Honda Accord, and Honda Civic. You would additionally add its domains like honda.com, international.honda and honda.com.au :

This can be a restricted instance to offer you an concept of what you may add on your model and opponents. Take the time to fill these out solely when you’re looking for a complete evaluation of every model.

When you’re executed, save the report so it turns into a preset you may return to at any time to check adjustments in AI visibility:

It’s going to then present up on the principle display screen of Model Radar alongside every other saved experiences that you would be able to all the time revisit sooner or later:

This completes the setup, and now you can begin benchmarking and analyzing the information.

3. Benchmark every model’s AI visibility 

To start out benchmarking, take a look at the Overview report in Model Radar. Pay attention to how metrics like your model’s mentions, citations, impressions, and Share of Voice change throughout numerous AI search platforms:

• Mentions: The variety of occasions your model is talked about in AI responses.
• Citations: The variety of occasions your model’s web site is cited as a supply in AI responses.
• Impressions: An estimation of publicity based mostly on how typically responses containing your model are proven in search outcomes.
• AI Share of Voice: How typically your model is talked about in AI responses in comparison with opponents.

You may also immediately evaluate your model to opponents for every AI search platform:

And likewise for every metric:

Maintain a file of all these benchmark metrics so you may evaluate them in opposition to one another sooner or later.

4. Take a look at how AI mentions you in comparison with opponents 

When you’ve benchmarked the fundamental visibility metrics, it’s time to look deeper. In contrast to conventional web optimization metrics, AI visibility extends past measuring if you seem. How you present up is equally essential.

Open the AI Responses report in Model Radar to see how every of the manufacturers in your evaluation is talked about in AI-generated solutions.

Begin by hovering over one of many manufacturers and choosing the “Solely model” choice:

You may also choose the opposite choices relying on what kind of research you wish to do:

• Solely model: This report will present you queries and AI solutions the place the one model talked about is the one you’ve chosen. It’s nice for assessing brand-specific queries and responses.
• With others: This report reveals queries and/or responses that include all of the manufacturers you’re evaluating. It’s nice for assessing matters which might be more likely to point out you alongside opponents.
• Others solely: This report reveals queries that solely point out the opposite manufacturers you’re evaluating in opposition to. It’s nice for locating matters that point out your opponents however not you.

Search for widespread language patterns for every model and the way AI differentiates one model from one other:

As an example, when evaluating Toyota and Honda, you may discover patterns, akin to Honda is usually talked about first, with attributes like affordability or compact vehicles being the most well-liked. Or that Toyota might seem second, framed round reliability and SUVs.

These variations improve your evaluation by offering a deeper understanding past top-level visibility. They present you (and stakeholders) every model’s positioning and perceived authority in AI solutions.

As you evaluate these responses, take note of parts like:

• Quantity of Mentions: Are you talked about roughly typically than your opponents throughout the matters your viewers searches for?
• Placement Inside the Response: Do you seem earlier than or after opponents? Which model will get the lion’s share of visibility within the response, higher capturing a searcher’s consideration?
• Depth of Protection: Are you summarized in a single line or described with detailed context like options, use instances, and advantages?
• Positioning and Framing: How does AI describe you (because the chief, an alternate, or simply an alternative choice)?
• Sentiment: Does the language round your model really feel constructive, impartial, or damaging?
• Credibility: Does the response embody robust credibility cues round your model or a competitor’s (e.g., “industry-leading,” “award-winning”)?

The sort of evaluation provides you the qualitative layer behind the numbers, exhibiting how AI fashions body, prioritize, and describe every model.

5. Analyze AI Share of Voice on your core matters 

You’ll be able to go even deeper in your evaluation by evaluating every model’s subject possession.

Take a second to record out the matters or product classes that you want to your model to be seen as a frontrunner for. After getting an inventory, add them to the Market or area of interest search field in Model Radar:

It’s simpler to go looking these one at a time.

For every subject, take a look at the Overview > AI Share of voice metrics:

It’s price benchmarking your AI Share of Voice for every of the matters you’d like to steer and the way you evaluate in opposition to different manufacturers in AI search.

You may also observe adjustments over time:

You may also take a look at the Subjects report to check every model aspect by aspect for widespread themes and sub-topics:

For instance, you might uncover insights like:

• Irrelevant matters to exclude with filters. For instance, queries about storage doorways could also be related to the subject of SUVs, however it’s unlikely any automotive model will earn significant mentions in these responses.
• Different opponents which might be linked to the subject. For instance, searchers incessantly join manufacturers like Honda, Lexus, Ford, and Tesla to SUVs. These could also be extra opponents to maintain an eye fixed on.
• Unbranded matters to enhance your visibility for. For instance, Toyota is talked about in additional responses about “finest SUVs,” whereas Honda is talked about extra for “compact SUVs”.
• Particular attributes linked to the subject. For instance, Honda might optimize its related merchandise to extend visibility for “MPG SUV” or “SUV for towing” queries.

6. Discover top-cited pages on every web site and shut gaps in your content material 

To search out every model’s top-cited pages in AI search, take a look at the Cited Pages report and apply the area filter proper above the record of pages:

Add your model first. Then examine every competitor one by one with the intention to extra simply see every model’s finest pages.

In case you’re doing an in depth evaluation, add every of the opponents’ web sites that you simply scoped out in Step 2 of this course of to see citations throughout all of them.

You may also check out completely different filter settings to concentrate on probably the most related pages or broaden the scope to ensure you haven’t missed something.

As an example, I like to make use of the Area scope setting, which mechanically contains the entered area and its subdomains. I additionally change the and/or toggle to “OR” to broaden the outcomes of this evaluation.

Relying on how deep you wish to go, it’s also possible to modify the filters on the high. For instance, say we wished to see Honda’s high citations for unbranded queries solely, we might apply this filter:

When you’ve added all of the filters you need, you’ll be left with an inventory of the top-cited pages in AI responses for every model.

Collect this record for every AI search platform you care about getting visibility for (like AI Overviews, ChatGPT and so on).

Then, repeat for every other opponents.

When analyzing these lists, search for content material alternatives and gaps you may shut in your web site. For instance, your content material might have:

• A visibility hole: Your model seems much less typically than opponents in AI citations.
• A subject hole: A competitor covers a subject, and also you don’t. Create a brand new web page to shut the hole.
• A format hole: AI typically cites a specific kind of content material format (e.g., guides, movies, evaluations) that you simply’re lacking.
• A freshness hole: A competitor wrote in regards to the subject extra just lately and will get extra visibility in consequence.

To shut these gaps, you’ll doubtless must create new content material or replace current content material. Ahrefs’ AI Content material Helper can assist with this. It analyzes your content material as you kind, permitting you to cowl important gaps.

7. Evaluate every model’s mentions across the net 

The ultimate examine price doing is to research how your opponents are talked about across the net. Model mentions in third-party content material are a robust issue that influences visibility in AI responses.

As a rule, AI will cite another person’s content material (even for queries about your model) than yours.

So, it’s price checking which sources affect details about your model and opponents. In Model Radar, you may take a look at the Net pages report to seek out all mentions of any model on-line.

To search out gaps in net mentions, I like to go looking every competitor individually in the principle search bar.

Then I exploit the filters within the Net pages report back to take away mentions on their very own web sites after which export the outcomes.

As an example, you should use “-site:yourcompetitor.com” to take away a competitor’s model mentions on their very own web site.

When you’ve eliminated irrelevant outcomes, export the record so you may evaluate it in opposition to your model and different opponents in a spreadsheet. Check out this easy template. Add your exported information to columns B–H, making certain they’re aligned appropriately. Then manually fill within the model in column A.

Maintain including every model’s export to the underside of the record.

When you’ve added information for a couple of manufacturers, right here’s what the comparability view will look like:

The empty cells level to missed alternatives. Begin by scanning for posts or responses that spotlight a number of opponents however go away your model out; these are doubtless your quickest wins.

Subsequent, analyze how your opponents earned their most impactful mentions:

• Content material varieties: Are their mentions embedded in weblog posts, analysis experiences, or case research? These codecs is likely to be price replicating or concentrating on in your outreach efforts.
• Publications and sources: Are the identical authors or web sites referencing them persistently? If that’s the case, they could be trusted voices in your area and appropriate candidates for outreach or collaboration.
• Marketing campaign triggers: Was the point out related to a product launch, viral submit, or a singular framework? Determine what sparked the eye to information your individual strategic strikes.
• Distribution techniques: Visitor articles, joint content material, or companion campaigns is likely to be amplifying their presence.

Reverse engineering what’s working on your opponents helps you shut visibility gaps, perceive which techniques truly transfer the needle in your class, and how you can make them work on your model.

8. Share your insights with key stakeholders 

By this level, your competitor evaluation will reveal how every model performs throughout AI search; who dominates visibility, who’s catching up, and the place you have got room to maneuver. The ultimate step is packaging these findings right into a story your crew can act on.

Begin by translating your observations into clear, actionable priorities:

• Repair: Shut visibility or notion gaps in comparison with opponents. These value you model authority in AI responses, typically attributable to inaccurate descriptions, lacking model information, or underperforming matters.
• Construct: Strengthen content material and topical protection the place opponents lead. Create or replace pages that align with the codecs and attributes AI rewards most.
• Affect: Strengthen relationships with publications, creators, and information sources that incessantly feed AI solutions in your class and point out opponents greater than you.

Then, summarize your findings in a stakeholder-ready report. Obtain our AI Search Competitor Evaluation Report template when you don’t have already got one.

Substitute all of the pink sections and summarize your key findings in scorecards and fast snapshots of key factors.

Additionally, it’s price making competitor monitoring a recurring workflow.

Run the identical evaluation month-to-month or quarterly to measure how share of voice and model framing evolve throughout AI fashions and markets.

AI search adjustments quick, and common benchmarking turns your one-off research into an ongoing strategic benefit.

Last ideas