The AI visibility gap hiding inside your best content

When AI Cites Your Content But Ignores Your Brand

In an AI visibility audit we ran this year for a national insurer, we put 40 questions its buyers actually ask into ChatGPT, Gemini, Perplexity and Google AI Overviews. That came to about 640 AI answers.

Then we looked at what was being cited and what was being named, two different measurements. Here’s what we found:

• Being cited: On this measure, the insurer looked excellent. They were cited as a source in about 148 of those answers, roughly 1 in 4. Few companies in its category were cited more often.
• Being named: The brand barely existed on this measure. It was named in only 16 answers, about 1 in 40. Six of those came from one minor product line. It was named zero times on its two biggest product categories.

The audit revealed that buyers were getting answers built on the insurer’s own content, with someone else’s name on them.

AI answer engines treat being cited and being named separately. The gap between them can be measured and worked on. If AI is already citing your pages, here is why it may still never name you, and what to do about it.

Key takeaways

• AI engines treat citing your page and naming your brand as two separate things. A national insurer we audited was cited in about 1 in 4 AI answers but named in only about 1 in 40.
• Most AI visibility dashboards blend mentions and citations into one score, which hides the gap. Tools can also miss shorter versions of your brand name.
• Engines appear to use brand pages to explain concepts, and other sources, such as comparison and review sites, to decide which brands to recommend.
• Visible, credentialed authors, a presence on the category sites engines already cite, and content that answers buyer decision questions all help tie your facts to your name.
• You can measure the gap yourself with 20 buyer questions run across ChatGPT, Gemini, Perplexity and Google AI Overviews, counting named and cited separately.

Why do most teams miss this gap?

Most AI visibility dashboards blend mentions and citations into a single score. A brand that is cited often but rarely named still ends up with a respectable number, and the gap disappears inside the average. On a blended score, the insurer above would look healthy. The problem only shows up when the two counts sit side by side.

Tools also undercount brand mentions in AI answers. In one of our runs, a tool counted three mentions of a brand where the answers actually showed seven. The engines had used a shorter version of the company name, and the tool wasn’t matching it. If your monitoring only looks for your full brand name, you are probably undercounting too.

The gap can also run the other way. A direct bank we audited was named in about 39% of answers, making it one of the two most visible brands in its market. Yet its own site ranked near the bottom of the bank sites the engines cited. The engines clearly knew the brand. They just weren’t learning about it from the bank.

Neither pattern shows up in a blended metric. Both become obvious the moment you count separately.

Why does the engine use your page but skip your name?

It’s not exactly clear how an engine decides what to cite and whom to name. Based on patterns we see across audits, here are four possible reasons why:

1. Your page works as a reference for AI answer engines

Educational content answers “what is” questions: what a deductible is, how a high-yield savings account works, and what term life insurance covers. AI engines rely on this material to explain concepts, and financial services companies publish a lot of it.

But when a buyer asks, “Which insurer should I use?” the engine is doing a different job. It is offering recommendations, and it appears to pull them from different sources than it uses for definitions. As a reference, your page provides the explanation. But someone else’s page supplies the names.

2. Your name isn’t in the passage

Engines are selective. They tend to lift content from a specific passage rather than the entire page. The paragraph an engine quotes usually has the fact it needs, but often not your company’s name. There’s no visible author, no credentials, nothing in that passage tying the insight back to you.

In regulated industries, content often goes out under a generic byline to keep review simple. One regional bank we looked at had about 150 published articles. Fewer than five had a named author, and none showed author credentials. To an engine, that content is useful but anonymous.

3. Your brand is missing from where engines pick names

When engines name brands, they seem to lean on places built for comparison: review platforms, category sites and aggregators. If your brand isn’t well represented there, your own site can be as authoritative as you like and still not get you named. For the AI engine, your page is a good source to cite for the facts. But it goes elsewhere for the shortlist.

4. Your content does not answer buying questions

Much financial content targets broad, high-volume topics. But buyers seek specific answers to their buying questions. Asking narrower questions, such as which policy makes sense for a family adding a teenage driver, or which bank suits a small business that handles a lot of cash deposits, helps them close in on a buying decision. If your content never answers those buying questions directly, the engine has little reason to put your name in the answer.

3 strategies to build content that carries your brand

The temptation to add more content is always there, and scientifically, it makes sense. A Nature study shows that when trying to improve a situation, our brains naturally choose an additive solution most of the time. But in marketing, fighting the urge to create more content is essential. What carries your brand is clear content that helps your brand show up where recommendations get made.

For a content strategy in regulated industries, that comes down to three things.

1. Credentialed authors, visible on the page. Real names, relevant credentials and a clear link to the company. Google’s guidance on helpful, reliable content for creators asks whether pages carry a byline and whether it leads to background information about the author. Google says it considers expertise and trust more important for topics that can affect a reader’s financial stability. Most financial services firms already have credentialed experts in-house. Their names just aren’t on the content.
2. Presence on the comparison and category sites the engines already cite. McKinsey found that in many cases a brand’s own sites make up only 5% to 10% of the sources AI search uses. In financial services, over 65% of those sources are publishers, user-generated content, and affiliate sites. Find the sites the engines cite when they recommend brands in your category, and make sure you appear there, accurately and in a way your compliance team has approved.
3. Content that answers buyer decision questions in the buyer’s own words. Not keyword phrases, but the questions your sales and service teams hear every week. Answer them directly, with your brand and your expert named in the same passage as the answer.

Treat this as the goal rather than a finished formula. What matters is that each piece is within a content team’s control.

How to measure it yourself

All you need to find out whether you have this gap is a spreadsheet and a few focused hours. Here are key steps you can take:

1. Write 20 questions in your buyers’ language. Focus on the questions people ask right before they choose, not broad educational topics.
2. Run them across the major engines. Ask each question on ChatGPT, Gemini and Perplexity about five times each, since answers vary from run to run, and check Google AI Overviews too. That gives you roughly 320 answers, half the scale as the insurer audit.
3. Count named and cited separately. For each answer, record whether your brand is named and whether your site is cited. Never average the two. Include shorter and informal versions of your name when you count.
4. Note who got named and which sites got cited instead. That is your to-do list: the competitors the engines prefer to recommend, and the comparison and review sites they rely on to do it.
5. Re-test every 60 to 90 days. Engines change, and so will your content. Track both numbers over time.

If your cited rate is strong but your named rate is low, you have the insurer’s problem. If your named rate is strong but your cited rate is low, you have the direct bank’s. Either way, you’ll know which problem you’re solving.

The engines already trust your content

If AI is citing your pages, the hardest part is done. The engines have decided your content is accurate and useful enough to build answers on. In categories where accuracy is regulated, that trust is worth a lot.

The work now is making your brand as visible as your facts. Put names and credentials on the page. Show up where recommendations get made. Answer the questions buyers ask when they’re ready to choose.

If you’d like to see how often your brand is cited versus named, and where the gap comes from, book Contently’s AI visibility audit. We measure both, separately.

 

Frequently asked questions

What is the difference between being cited and being named in an AI answer?

Being cited means the engine used your page as a source, usually shown as a link or reference. Being named means the answer mentions your brand in its text, often as a recommendation. An engine can do one without the other, so each needs its own count.

Why does AI cite my content but not mention my brand?

Based on our audits, the most likely reasons are that your pages work as reference material rather than recommendations, the passages engines quote don’t carry your brand or an author, your brand is thin on comparison and review sites, and your content answers broad questions instead of the specific ones buyers ask before they choose.

Can my AI visibility dashboard show this gap?

Only if it reports mentions and citations separately. Many tools blend them into a single score, where a brand that is cited often but rarely named can still look healthy. It’s also worth checking whether your tool matches shorter or informal versions of your brand name.

How often should we measure AI citations and brand mentions?

Every 60 to 90 days. Engines change, and so does your content, so tracking both numbers over time shows whether your work is closing the gap.