How to compare backlinks for brands cited in AI answers with Ahrefs

You compare backlinks for brands cited in AI answers in two stages: first you find out which brands the answers cite, and then you put those brands' link profiles next to yours. The first stage happens outside your usual SEO reports. You ask ChatGPT, Gemini, Perplexity, and Claude the questions your buyers ask, such as "What's the best scheduling software for a small dental practice?", and write down which brands each answer names and which pages it cites as sources. Ahrefs sells an add-on called Brand Radar that tracks AI mentions and cited pages across a large set of prompts, which saves you from asking every question by hand.

The second stage is where Ahrefs' link index does its work. In Site Explorer, you enter a domain and open its Referring domains report to see the distinct websites that link to it, or its Backlinks report to see each individual link with the page it sits on and its anchor text, which is the clickable wording of the link. Batch Analysis lets you paste in a list of domains and compare their Domain Rating and referring-domain counts side by side. Link Intersect takes several competitor domains plus your own and returns the sites that link to the competitors but not to you.

Those tools hand you the numbers, but they don't tell you what the numbers mean for AI answers. A brand can lead you by thousands of backlinks because one partner site links to it from every page footer, and that tells you very little. Another brand can have a link profile much like yours and still get cited every time, because it published the clearest comparison page in your category. Link Intersect returns every site that matches, including directories, scraped copies, and sites in unrelated industries, so the list needs sorting before it's useful.

A good result answers three questions:

• Do links explain the gap? For each cited brand, the comparison should say whether its link profile is meaningfully stronger than yours or whether something else, usually content, explains why it gets cited.
• Which linking sites are worth pursuing? The list should hold real publications, review sites, and communities in your category that already vouch for brands like yours.
• What should you do next? Each gap should come with a move: a site to pitch, a page to build, or a third-party roundup to get into.

Kite does that work for you. It finds the cited brands with its own tracking of the four engines, reads their link profiles through your Ahrefs account, and turns the comparison into a short list of credible gaps and a plan to close them.

Start by asking Kite in Slack, in your own words:

"Find out why ChatGPT and Perplexity keep recommending our competitors, and use Ahrefs to see whether their backlinks explain it."

Kite is an AI marketer in your Slack. It researches your business, builds a growth strategy, and ships the work that brings you new customers. Because it already knows your category, your buyers, and your competitors from that research, it proposes the buyer questions to track and the brands to watch without a long brief. Confirm the questions or add your own, and Kite starts.

You'll need to connect your own Ahrefs account through OAuth. Kite only reads from Ahrefs, and each request it makes draws on your Ahrefs plan's API allowance. If a task needs Ahrefs before it's connected, Kite sends you a link in Slack to connect it. The AI-answer tracking is built into Kite, so you don't need Brand Radar or any other AI-visibility subscription for this job.

Reading link data and comparing it is research, so Kite does it and reports back. Outreach and new pages come later in the work, and Kite brings those to you for approval before anything is sent or published.

Kite starts by finding the brands that AI answers actually cite for your buyers' questions, because those are the brands worth comparing. Your usual list of SEO competitors is a different list. The brands that outrank you on Google aren't always the ones ChatGPT recommends, and a small competitor can show up in AI answers well ahead of a bigger one.

Kite runs each question through ChatGPT, Gemini, Perplexity, and Claude. For every answer, it records whether your brand appears, which competing brands appear, and which pages and domains the answer cites. It saves the answers and their citations and measures each brand's share of voice across the full set of questions.

From those results, Kite picks the brands to compare. It favors brands that appear across several questions and more than one engine, since a brand cited once by one engine tells you little. It also notes when the cited page belongs to someone else. AI answers often cite a third-party roundup, a review site, or a forum thread that names a brand, and that changes what the backlink comparison needs to look at.

Kite compares link profiles by counting referring domains, meaning the distinct websites that link to a brand, rather than raw backlinks. A hundred links from one site count as one vote, which keeps sitewide footer links and partner badges from skewing the comparison.

Through your Ahrefs account, Kite reads the referring domains for each cited brand's domain and for your own. It asks Ahrefs for the columns the comparison needs, such as each linking site's Domain Rating and the date Ahrefs first saw the link. That lets Kite compare more than size. It sees how many of each brand's linking sites carry real authority, and whether a brand has been gaining links quickly in recent months, which often lines up with a recent rise in AI answers.

Kite reads each domain as a whole, with its subdomains, so a brand's documentation site or community forum counts toward its profile. When one section of a site matters more, such as a competitor's /blog/ folder, Kite reads only the pages under that path.

Kite decides, brand by brand, whether the citation gap is a link gap or a content gap, because the fix for each one is different. Links are one reason a brand gets cited, and Kite weighs them alongside what the cited pages actually say.

For each cited brand, Kite reads the pages the answers cited and checks them against the link comparison. When a competitor has many more relevant referring domains than you do, links likely explain much of the gap, and the work is earning coverage on the sites that vouch for that competitor. When a competitor's link profile looks much like yours, the reason usually sits on the cited page, such as a plain comparison table, a direct answer to the buyer's question, or pricing that's easy to find. Then the work is building a better page on your own site.

Kite also reads the backlinks pointing at the exact cited page, not just at the brand's domain. A cited page that many sites link to is easy for search engines and answer engines to find, which helps explain why it keeps showing up. When the cited page is a third-party roundup that names your competitors, Kite treats getting your product onto that roundup as its own goal.

Kite keeps the linking sites that cover your category and link to two or more cited brands without linking to you, because those sites have already shown they'll point readers at products like yours. A site that links to every brand in your market except yours is the strongest prospect on the list.

To build that list, Kite asks Ahrefs for one backlink from each referring domain for every cited brand, choosing the strongest link from each site. That turns thousands of links into a clean list of websites, each with the page that carries the link, that page's title, and the anchor text. Kite compares those lists against your own referring domains and removes every site that already links to you.

Then Kite sorts what's left. It drops directories, scraped copies of other sites, link farms, and sites in unrelated industries. It reads the linking page's title and anchor text to see what kind of mention each link is, whether that's a review, a "best tools" roundup, an integration directory, a guest article, or a customer story. It ranks the remaining sites by how many cited brands each one links to, how much authority it carries, and how closely its readers match your buyers.

Kite reads the individual backlinks pointing at each cited brand to learn which of its pages attract links and how other sites describe them. Each backlink comes with the linking page, its title, and the anchor text the linking site chose.

This shows what your market likes to cite. If most of a competitor's links point at a free pricing calculator, and the anchor text reads like "this calculator" or "cost estimate tool," that calculator is doing much of the work for that brand. Kite notes which of those assets you could answer with a stronger version of your own, since a page that earns links on its own helps both your search rankings and your chance of being found by the engines.

When the comparison is done, Kite posts a summary in the Slack thread where you asked. It names the brands it compared, says for each one whether the gap is mostly links or mostly content, and lists the best sites to pursue with the reason each one made the list. When the comparison needs more room than a Slack message, Kite publishes it as a report with sortable tables of every brand and every prospect site.

From there, Kite proposes the work that closes the gaps:

• Pitches. For each prospect site, Kite drafts a pitch that names the page your product would fit on and why it belongs there, and it sends them once you approve them in the thread.
• New pages. Where content explains the gap, Kite writes the comparison page, answer page, or linkable asset, and publishes it on a site Kite hosts or through your connected CMS after you approve it.
• Measurement. After the work ships, Kite reads your referring domains again to see which new sites link to you, and it runs the same buyer questions through the four engines to measure whether your brand now appears.

You can ask Kite to repeat the comparison on a schedule, such as the first Monday of every month, and post what changed in the same thread.

When Kite finishes, you have a clear answer to whether backlinks explain why AI answers cite your competitors, plus a list of the specific sites and pages that would change it.