What blog posts should you write to be mentioned in ChatGPT?

The rise of generative search and AI assistants is reshaping SEO and content strategy. If the primary goal used to be ranking highly in Google search results, marketers are now increasingly asking a different question: how can a brand or piece of content be featured in ChatGPT answers and other AI-generated responses?

A recent study analyzing query fan-out behavior in ChatGPT highlights an important trend: AI systems are far more likely to perform web searches for commercial queries than for informational ones.

This directly impacts the types of content most likely to appear in AI-generated answers.

What is query fan-out and why does it matter?

ChatGPT does not always generate answers solely from its training data. For some queries, the system performs additional web searches — a process known as query fan-out.

The model breaks the user’s request into multiple subtopics, runs parallel searches, and synthesizes the final answer from the retrieved sources.

In practice, this means AI systems search not only for the exact query but also for related subqueries such as:

• product comparisons;
• recommendations;
• feature evaluations;
• alternatives;
• rankings;
• decision-support content.

If a company’s content does not cover these “branches” of search intent, the likelihood of appearing in ChatGPT responses decreases significantly.

Key finding: ChatGPT favors commercial content

The study analyzed 90 prompts across three industries:

• beauty;
• legaltech/regtech;
• IT/tech.

The results revealed a clear pattern:

• commercial prompts triggered fan-out in 78.3% of cases;
• informational prompts did so in only 3.1%.

Out of 20 prompts that triggered fan-out:

• 18 were commercial;
• only 2 were informational.

Moreover, even when the original query was informational, ChatGPT often transformed it into a commercial or evaluative search path.

For example:

• AI tools queries evolved into searches for specific solutions;
• general questions about document management systems turned into product comparisons;
• broad skincare prompts shifted toward product recommendations and rankings.

What this means for content strategy

The main takeaway for marketers is not that informational content is no longer valuable.

The issue is that informational content alone is unlikely to secure visibility in AI-generated answers.

If a brand wants to be surfaced in ChatGPT responses, it needs content aligned with decision-making behavior — content that helps users compare, evaluate, and select solutions.

Which types of content have the highest potential for AI visibility?

Best-Of and Shortlist Articles

Content formats such as:

• “Best CRM Platforms for E-commerce”;
• “Top AI Tools for Marketing Teams”;
• “Best Automation Platforms for Digital Advertising”.

These articles align naturally with fan-out logic because AI systems frequently generate shortlist-oriented searches.

Comparison Pages

Content formats including:

• “HubSpot vs Salesforce”;
• “GA4 vs Adobe Analytics”;
• “Asana vs Monday: Which One Should You Choose?”

Such pages directly match evaluative intent, which ChatGPT commonly generates during query expansion.

Alternatives Pages

AI systems regularly search for alternatives to products and services.

Examples include:

• “Alternatives to Notion”;
• “Ahrefs Alternatives”;
• “Universal Analytics Replacements”.

For B2B companies, this can become one of the most effective formats for AI visibility.

Feature-Led Content

Articles built around specific features or use cases:

• “CRM Platforms With AI Analytics”;
• “Tools With Multi-Touch Attribution”;
• “Customer Support Automation Platforms”.

This type of content helps AI systems connect user needs with relevant solutions.

Evaluation FAQs

Pages answering decision-oriented questions are also becoming increasingly important:

• How do you choose the right platform?
• Which evaluation criteria matter most?
• What type of business is the solution best suited for?
• What limitations does the product have?

These formats create additional entry points for AI retrieval systems.

Why ToFU content can no longer exist in isolation

Traditionally, content marketing relied heavily on top-of-funnel educational articles such as:

• “What Is a CDP?”;
• “What Is Performance Marketing?”;
• “How Does SEO Work?”

However, the study suggests that if an article only explains concepts without including evaluative or recommendation-oriented elements, its chances of appearing in AI-generated answers decrease.

Today, the most effective strategy is ToFU content combined with commercial bridges.

For example, even educational articles should include:

• platform examples;
• use cases;
• selection criteria;
• solution comparisons;
• recommendation sections.

These are precisely the types of elements ChatGPT may use during query fan-out expansion.

How to adapt your content model for AI search

For mid-sized and enterprise businesses, this means rethinking content architecture entirely.

An effective AI-focused content strategy should include:

Commercial Content Clusters

Around each core product or category, brands should build:

• comparison pages;
• alternatives pages;
• buyer guides;
• evaluation content;
• integration and use-case content.

Coverage Across the Entire Decision Journey

AI increasingly acts as an assistant during the buying and evaluation process. As a result, brands need content covering:

• awareness;
• consideration;
• evaluation;
• selection.

Recommendation and Comparison Language

Content should naturally include phrases such as:

• “best for…”;
• “ideal for companies that…”;
• “alternative to…”;
• “worth choosing if…”.

These semantic patterns are commonly used by AI systems when generating recommendations.

Limitations of the study

At the same time, the findings should not be treated as universal rules.

The researchers emphasize that:

• the sample size was limited;
• informational prompts dominated the dataset;
• results may vary across industries;
• this was not an analysis of ChatGPT’s internal architecture, but rather an observation of system behavior.

Nevertheless, the research demonstrates a clear trend: AI systems increasingly prioritize content that supports decision-making.

Conclusion

AI-powered search is transforming not only SEO but the entire logic of content marketing.

To appear in ChatGPT answers, publishing informational articles alone is no longer enough. Generative AI systems increasingly prioritize content that:

• compares;
• recommends;
• evaluates;
• helps users make decisions.

For businesses, this means shifting from purely educational SEO toward hybrid content that combines expertise with commercial usefulness.

This approach is likely to offer the greatest visibility potential in the emerging AI-driven search ecosystem.

Read this article in Ukrainian.