What happened
Minuttia, an SEO and AEO agency, published a study analyzing how AI assistants and Google Search recommend software. The research found that a significant portion of the 'best tools' and comparison lists cited by these platforms are self-promotional content written by the software vendors themselves.
Minuttia published the Self-Promotional Content Data Study, its second large-scale analysis of how AI search engines answer software-related questions. The study examined 18,084 questions across six platforms: Google Search, AI Overviews, ChatGPT, Gemini, Claude, and . The questions were categorized into three types: 'best alternatives to [product],' 'how do [product] and [competitor] compare,' and 'best tools for [category].'
The analysis covered 5,983 software brands across 196 categories. Researchers read and labeled every page cited by the platforms, totaling 311,766 unique pages cited 686,749 times. The primary focus was on identifying whether the cited listicles were written by the software vendors themselves and whether they disclosed this authorship.
The study found that 34% to 47% of the listicles cited for 'alternatives' questions were vendor-written. For 'X vs Y' comparison questions, this figure rose to 56% to 65%. Furthermore, 87% of these vendor-written lists ranked the vendor's own product first, and only 16% of the pages disclosed who wrote them.
The results varied by platform. ChatGPT tended to cite the vendor's own website, while rewarded pages that ranked well and were useful to read. The share of vendor lists ranged from 12% to 76% depending on the specific software category. Minuttia noted that the self-promotional shares are a floor, as pages that could not be read were counted as not self-promotional.
The study also highlighted the volatility of AI search results. When Minuttia re-asked 2,000 questions to in September, only 28% of the pages listed were the same as those in the August answers. The full study, methodology, and underlying data are available for public review.
Source details: techbullion.com ↗
Why it matters
This finding highlights a critical transparency gap in AI-driven search results. When AI assistants recommend software, users often assume the sources are neutral third-party reviews. However, if the majority of these sources are vendor-authored and undisclosed, the recommendations may be biased toward the vendor's own product. This undermines user trust in AI search engines and complicates the software selection process for businesses and individuals relying on these tools for objective guidance.
The study reveals a significant conflict of interest in AI-generated software recommendations. Users asking AI assistants for software advice are often receiving recommendations based on content written by the very companies selling the software. This creates a where the vendor's product is likely to be ranked first, regardless of its actual merit relative to competitors.
The lack of disclosure is a major concern. With only 16% of vendor-written lists indicating their authorship, users are unlikely to know that the 'objective' comparison they are reading is actually a marketing asset. This erodes the perceived neutrality of AI search engines and can lead to poor purchasing decisions.
For businesses, this means that relying solely on AI assistants for competitive analysis or software selection may yield skewed results. Companies may need to implement additional verification steps or seek out independent reviews to ensure they are making informed decisions.
The findings also have implications for the AI industry. As AI assistants become more integrated into daily workflows, the quality and integrity of the sources they cite become increasingly important. This study provides concrete evidence that current sourcing mechanisms may be insufficient to prevent biased or self-promotional content from influencing user decisions.
Interactive Mechanism: How It Actually Works
Explore the underlying technology behind this development interactively.
A route planner searches possible journeys using explicit rules. What does this illustrate about AI?
What to watch next
Watch for regulatory or platform-level responses to address the lack of disclosure in AI-cited sources. Additionally, monitor whether AI platforms begin to explicitly label vendor-written content in their search results to mitigate and improve transparency for users.
Monitor whether AI platforms like ChatGPT, Gemini, and introduce new features to label or flag vendor-written content in their search results. Such transparency measures could help users identify potential biases in the recommendations they receive.
Watch for regulatory developments regarding AI search engines and the sources they cite. If the lack of disclosure in AI recommendations is deemed misleading, there may be calls for stricter regulations on how AI platforms source and present information.
Observe how software vendors and SEO agencies adapt their strategies in response to these findings. Vendors may face pressure to be more transparent about their content, while SEO agencies may develop new techniques to help clients achieve more neutral visibility in AI search results.
Look for follow-up studies or independent analyses that verify Minuttia's findings. Independent replication of the study would strengthen the evidence base and potentially lead to broader industry consensus on the extent of self-promotional content in AI search results.