ResearchJanuary 20, 2026

We Analyzed 10,000 AI Recommendations: Here's What Gets Brands Mentioned

We Analyzed 10,000 AI Recommendations: Here's What Gets Brands Mentioned

What determines which brands AI recommends? We analyzed 10,000 AI recommendation responses across ChatGPT, Perplexity, and Claude, spanning 50 product categories.

Here's what we found.

Finding #1: Big brands don't always win

In 34% of queries, lesser-known brands outranked industry giants in the top 3. Some examples:

Smaller brands win when their positioning specifically matches the query, when review sentiment is strongly positive, when recent coverage is favorable, and when community advocacy is high.

Takeaway: Targeted positioning and genuine advocacy can overcome brand size disadvantages.

Finding #2: Position #1 gets 3x more consideration

We surveyed 500 users who received AI recommendations. Consideration rates by position:

The first brand mentioned gets 3x more consideration than the third, and 7x more than brands mentioned fourth or later. In traditional search, ranking #5 still gets clicks. In AI, the drop-off is brutal.

What influences position: direct relevance to the query, recent press coverage, high review scores, and clear market positioning. Generalists rank lower than specialists.

Finding #3: Different platforms give different answers

67% of queries returned different top-3 recommendations across ChatGPT, Perplexity, and Claude.

How often do platforms agree on #1?

Example: "Best help desk software for startups" - ChatGPT: Intercom #1, Freshdesk #2, Help Scout #3 - Perplexity: Zendesk #1, Intercom #2, Freshdesk #3 - Claude: Help Scout #1, Intercom #2, Zendesk #3

Same query, three different top picks. Each platform has different training data, different recency biases, and different weighting algorithms.

Takeaway: Tracking only one platform gives an incomplete picture.

Finding #4: Negative mentions are worse than no mention

How mention type affects buyer consideration:

Being mentioned negatively drops consideration below not being mentioned at all. Common triggers for negative mentions: pricing complaints, usability issues, customer service problems, and negative press coverage.

Takeaway: Monitor sentiment, not just presence. A negative mention strategy might be to stay out of certain conversations until underlying issues are addressed.

Finding #5: Fresh content improves ranking

AI platforms with real-time search (ChatGPT, Perplexity) show clear recency bias.

Position change based on content freshness: - Content in last 30 days: +0.8 positions - 30-90 days: No significant change - No content in 90+ days: -1.2 positions

Brands publishing regularly rank an average of 2 positions higher than those with stale content. What counts most: news coverage and PR, blog posts with original data, product announcements, and recent review activity.

Takeaway: Your content calendar directly impacts AI visibility.

Methodology

What to do with this

  1. Don't concede to larger competitors - 34% of top recommendations go to non-leaders
  2. Obsess over position, not just presence - The gap between #1 and #3 is 3x consideration
  3. Track all major platforms - 67% disagreement means single-platform tracking is incomplete
  4. Treat negative mentions as emergencies - Worse than being absent
  5. Publish consistently - Fresh content = 2 positions higher on average

OakData monitors your brand across every major AI platform - tracking position, sentiment, and competitor movements over time.