AI Recommendation Rates Have Shifted Consumer Loyalty
New research shows AI models often omit brand recommendations, complicating how companies capture intent.
Updated on Sept. 28, 2026 in Artificial Intelligence
Live Poll
Do you trust AI tools to provide neutral and unbiased recommendations for products and services?
Leoprd published its Reputation to Revenue 2026 report, revealing that brands were absent from 36% of unbranded AI responses analyzed. The study examined 31,200 AI interactions to understand how large language models influence consumer decision support.
Why it matters
As 91% of shopping prompts involve research, the tendency for AI to rely on third-party sources threatens traditional brand loyalty. This shift forces companies to rethink how their presence in media and review data impacts AI-generated outcomes.
Analysis shows 86% of citations in AI responses originated from external sources rather than brand-owned properties. Furthermore, only 40% of repeated test groups returned consistent recommendation outcomes, highlighting significant instability in model results.
The players
Leoprd
A research firm focused on tracking digital brand performance and AI-driven consumer interaction patterns.
Prompt Cowboy
A data provider specializing in capturing and anonymizing large-scale user-AI prompt interactions.
The details
Leoprd analyzed 31,200 AI responses against a set of 214,000 monitored-brand responses using anonymized prompt data provided by Prompt Cowboy. The study found that AI recommendation outcomes are heavily influenced by patterns of evidence in external media coverage and reviews. This mechanism means that a brand's visibility is largely dependent on its third-party digital footprint rather than direct marketing efforts.
Timeline
September 2026: Leoprd published the Reputation to Revenue 2026 report.
October 13-14, 2026: Additional findings will be presented at the HumAIn conference.
The Tech Race
This study updates the industry understanding of how AI search models deviate from traditional web-based referral patterns. It highlights a departure from legacy SEO practices where brand-owned properties were the primary drivers of discovery.
Businesses relying on traditional customer acquisition must now address that 86% of AI citations come from external sources. Organizations should anticipate that AI-led research will increasingly replace manual web searches, shifting customer service needs away from self-service contact.
The takeaway
AI models are actively restructuring brand discovery by favoring external research over brand-owned content. Readers should monitor the upcoming HumAIn conference presentation to see if model recommendation consistency improves from its current 40% repeatability rate.
What happens next
Further findings from the Reputation to Revenue 2026 report are scheduled to be presented at the HumAIn conference on October 13-14, 2026.
Further reading
For broader insights on machine-generated consumer patterns, visit our Artificial Intelligence section.
More information
To review the full data set, access the Reputation to Revenue report via the publisher's portal.
Live Poll
Do you trust AI tools to provide neutral and unbiased recommendations for products and services?
