What Makes AI Recommend a Brand More Often? Insights from the RankCaster AI Study

AI is becoming part of the customer journey long before a potential client visits a company’s website.

People ask AI systems which agency to hire, which software to choose, which course to take, or which provider can solve a specific problem. The answer may include several brands, but not every company that has relevant content will be mentioned.

This creates a new marketing challenge:

> How can a brand increase the likelihood that AI systems will discover, understand, and recommend it?

At Terekhin Digital Crew, we reviewed a recent RankCaster AI study that explores this question using data from four brands and 8,058 observations.

You can read the original research here:

How to Increase Brand Mentions in AI: RankCaster AI Data from 8,058 Observations

From measuring visibility to influencing it

Many AI visibility tools are designed to answer a diagnostic question:

> Where is a brand already mentioned by AI?

This information is useful. It helps marketers see which prompts generate visibility, which competitors appear in answers, and which URLs AI systems use as sources.

However, monitoring alone does not explain what a marketing team should do next.

It does not necessarily answer:

• Which content gaps are limiting the brand?
• Why is a competitor recommended instead?
• Is the brand’s own website enough?
• Do external mentions influence AI answers?
• How many citations are associated with a meaningful increase in brand mentions?
• What should the team do when visibility starts to decline?

This is the distinction RankCaster AI makes in its positioning.

RankCaster AI describes itself as a proactive AI Visibility Marketing platform designed to help companies change how AI systems discover, understand, and recommend their brands.

Unlike tools that only measure visibility, the platform identifies gaps, analyzes why competitors are recommended, and helps marketing teams take action to increase brand presence in AI-generated answers.

What the study measured

From August to September 2026, RankCaster AI tracked four brands that published content through its Content Manager:

• a marketing agency in Israel;
• a mortgage broker in Israel;
• a contractor-matching platform in the United States;
• an education center in Poland.

The study produced 8,058 observations over six weeks.

Each observation represented one weekly check of one target prompt for one brand.

For every AI response, RankCaster AI recorded two separate events:

This distinction is critical.

A page can be cited without the brand being mentioned prominently. For example, AI may use a page as a source but describe the answer without naming the company.

A brand can also be mentioned because an AI system found it in an external source, such as a directory, partner website, social platform, or industry publication.

That is why citation rate and mention rate should be tracked separately.

The main finding: signals work better together

The study divided visibility signals into two categories.

Owned signals

Owned signals come from the brand’s own website. They include:

• blog articles;
• service pages;
• case studies;
• research;
• expert content;
• FAQs;
• analytical reports.

These pages help AI systems understand what the brand does, which services it offers, which markets it serves, and which topics are connected to its expertise.

External signals

External signals come from third-party sources. They include:

• social media;
• business directories;
• partner websites;
• industry platforms;
• media publications;
• aggregators;
• cross-brand content.

These sources provide context beyond the brand’s own claims. They help connect the brand with a category, service, market, or topic across multiple locations on the web.

The RankCaster AI study found the following relationship:

• no citation signals: 0.5% mention rate;
• owned content only: 35%;
• external signals only: 40%;
• owned and external signals combined: 58%.

The important point is not that external mentions replace a brand’s own content.

A brand still needs authoritative pages that clearly explain its services and expertise. But those pages may perform better when external sources reinforce the same associations.

From a digital marketing perspective, this is a shift from isolated content production to distributed brand presence.

Citation frequency and AI mentions

The study also grouped observations according to the number of relevant citations a brand received per week.

The results were:

• 0 citations: 0.5% mention rate;
• 1-5 citations: 21%;
• 6-20 citations: 50%;
• 21 or more citations: 61%.

The difference between no citations and 6-20 citations was 49.5 percentage points.

This does not mean that six citations automatically produce a 50% mention rate. AI visibility is influenced by many factors:

• topic competitiveness;
• prompt wording;
• content quality;
• source authority;
• language and location;
• indexing speed;
• competitor activity;
• relevance to the exact user intent.

The figures should therefore be treated as directional benchmarks, not as a guaranteed ranking formula.

The more useful conclusion is that relevant signals appear to accumulate. A brand that appears consistently across sources may be more likely to enter the context AI systems use when forming an answer.

What happened in the four projects?

The individual projects illustrate why AI visibility requires ongoing measurement.

Marketing agency in Israel

The agency recorded a 40.5% mention rate and an average of 8.4 citations from its own domain per week.

For one prompt, the brand appeared in 100% of answers when it received 37 citations per week. Later, when citations declined to 9-12 per week, the mention rate for the same prompt fell to 17-20%.

The agency also received 664 cross-brand citations from a partner in the same portfolio.

Mortgage broker in Israel

The mortgage broker recorded a 41.7% mention rate across 13 tracked prompts.

It received an average of 7.6 citations per week and maintained a relatively stable presence among AI sources.

The project also received 698 cross-brand citations.

Contractor-matching platform in the United States

This project provided the clearest before-and-after comparison.

After content publication began, the mention rate increased from 6.3% to 14.4%.

The number of citations increased from 0.5 to 2.5 per week.

That represented:

• an increase of 8.1 percentage points;
• a relative increase of 130%.

The result was positive, but 2.5 citations per week remained in the lowest citation range. More content or stronger external signals were needed to move the brand into the ranges associated with higher mention rates.

Education center in Poland

The first article was published on August 26.

Within three weeks, citations increased from 13 to 65 per week.

For the prompt “SMM courses in Europe”:

• 12 citations corresponded to a 50% mention rate;
• 18 citations corresponded to 67%.

For the prompt “SEO courses,” the mention rate fell to 0% after citations stopped.

This example highlights the importance of topical alignment. Content must address the same need represented by the monitored prompt. General visibility is not enough if the page does not answer the specific question.

Why visibility can decline

One of the most valuable findings concerns degradation.

In one case, a prompt related to a Russian-speaking event advertising agency continued to generate 10-17 citations per week. However, the brand’s mention rate fell from 100% to 43% over six weeks.

The brand was still being cited. So why did mentions decline?

The likely explanation was increased competitor activity around the same topic. AI continued to use the brand’s content, but other companies became more prominent in the final answers.

This is an important lesson for marketing teams:

> A citation does not guarantee a recommendation.

When mention rate falls despite stable citations, the team should:

• review which competitors have entered the answers;
• compare their content and positioning;
• identify missing information;
• update relevant pages;
• add new examples and proof points;
• strengthen external signals;
• monitor the same prompts again.

This turns degradation into an actionable competitive signal rather than a passive reporting metric.

What does this mean for digital agencies?

The study has several implications for agencies working on SEO, content marketing, PR, and brand visibility.

1. Content needs a defined visibility objective

Publishing articles without connecting them to target prompts makes it difficult to evaluate their impact.

A stronger process starts with questions such as:

• Which commercial prompts currently exclude the brand?
• Which competitors appear instead?
• What information does the current answer rely on?
• Does the brand have a page that directly addresses the topic?
• Which external sources could reinforce the association?

2. Owned content and external presence should be planned together

A content calendar should not focus only on articles published on the client’s own domain.

It should also consider:

• relevant directories;
• partner content;
• social profiles;
• industry platforms;
• expert commentary;
• media coverage;
• cross-brand references where genuinely relevant.

The goal is not to create artificial mentions. The goal is to make the brand consistently discoverable in the contexts that matter.

3. Portfolio relationships can create additional signals

The cross-brand results are especially relevant to agencies managing multiple clients.

When brands operate in related categories, natural and useful references between them may create additional external signals.

However, these references must be justified by the subject matter. They should help the audience understand a service, category, market, or solution, not simply create a link exchange.

4. AI visibility requires continuous optimization

The competitive environment changes. New pages are published, competitors become more visible, and AI systems may change the sources they use.

For that reason, AI visibility should be managed as a cycle:

Our view at Terekhin Digital Crew

The RankCaster AI study supports a broader shift in digital marketing.

SEO and AI visibility should not be treated as separate content activities. Both depend on whether a brand can be clearly understood, consistently associated with relevant topics, and supported by useful information across the web.

The difference is that AI visibility introduces a new measurement layer.

It is no longer enough to ask:

• Does the page rank?
• Does the website receive traffic?
• Is the brand mentioned somewhere online?

Marketing teams also need to ask:

• Is AI using the page as a source?
• Is the brand named in the answer?
• Which competitors are recommended instead?
• What sources appear around the topic?
• Is the brand’s visibility increasing or declining over time?

This is where a proactive platform such as RankCaster AI can complement the work of a digital agency.

The platform combines:

• AI visibility monitoring;
• prompt-level gap analysis;
• competitor analysis;
• content recommendations;
• content generation;
• publishing workflows;
• external source discovery;
• tracking of citations and brand mentions.

Its model has been trained on more than 5 million AI citations and is designed to analyze the factors that may affect signal strength, from the publishing source to the relevance and structure of the content.

Final takeaway

The RankCaster AI study analyzed 8,058 observations across four brands over six weeks.

The findings indicate that:

• no citation signals were associated with a 0.5% mention rate;
• owned content alone reached approximately 35%;
• external signals alone reached approximately 40%;
• the combination of both reached 58%;
• brands with 6-20 citations per week reached approximately 50%;
• brands with 21 or more citations reached approximately 61%.

The key lesson for brands and agencies is simple:

> AI visibility is not only about being cited. It is about becoming a recognizable and consistently supported answer within a relevant topic.

A brand’s own content creates the foundation. External sources reinforce the association. Monitoring reveals what is changing. Proactive optimization turns that information into action.

Read the original RankCaster AI research:

How to Increase Brand Mentions in AI: RankCaster AI Data from 8,058 Observations

Editorial note: the results discussed here come from a four-brand sample and should be treated as directional research findings, not as a universal guarantee. Terekhin Digital Crew is presenting the study from an agency perspective and is not the publisher of the original research.

Google’s current search guidance continues to emphasize helpful, reliable, people-first content and clear, descriptive page structure rather than mechanical optimization alone. developers.google