Recommendation Rate
The share of observed decision-intent answers in which an AI system includes a brand as a recommended option under a declared measurement protocol.
Defined at the home of Machine Relations: Recommendation Rate in the Machine Relations glossary
Recommendation Rate measures how often AI engines actively recommend your brand — not just mention it — when users ask purchase-intent questions in your category. It is the highest-signal metric in Machine Relations because a recommendation carries direct commercial intent.
Machine Relations publishes its own definition of this term in its glossary: Recommendation Rate, defined by Machine Relations, a related property.
Why Recommendation Rate Matters
Not all AI citations are equal. Being mentioned in a list of ten alternatives is categorically different from being the AI's top recommendation. When ChatGPT says "I recommend [Brand] for this use case," the brand is the pick, not one name in a list. Recommendation Rate isolates those answers from passive mentions.
AI engines typically recommend 2-3 brands per query. That is the entire addressable surface. If your Recommendation Rate is zero, you are not in the consideration set for AI-mediated purchases — regardless of your Google rankings or social following.
How to Measure Recommendation Rate
- Define 20-50 high-intent queries buyers ask in your category ("best X for Y," "which tool should I use for Z")
- Run each query across ChatGPT, Perplexity, Gemini, and AI Overviews
- Classify each brand appearance as either recommendation (active endorsement) or mention (passive inclusion)
- Calculate:
Recommendations / Total Queries = Recommendation Rate
A Recommendation Rate above 15% indicates strong category authority. Below 5% signals an urgent gap.
How to Increase Recommendation Rate
AI engines recommend brands they trust through multiple corroborating signals:
- Earned media density — Tier-1 publications citing your brand as a leader, not just a participant. 82-89% of AI citations trace back to earned media, the sources brands neither own nor pay for.
- Specificity — Content with concrete stats, named features, and differentiated positioning gives the AI something specific to recommend.
- Recency — Fresh coverage from the last 90 days weights heavily in recommendation decisions.
Pair Recommendation Rate with AI Mention Rate for the full picture: mention rate tells you if AI knows you exist, recommendation rate tells you if it trusts you enough to endorse. Both feed into your AI Visibility Score.
Frequently Asked Questions
Why Recommendation Rate Matters?
Not all AI citations are equal. Being mentioned in a list of ten alternatives is categorically different from being the AI's top recommendation. When ChatGPT says "I recommend [Brand] for this use case," the brand is the pick, not one name in a list. Recommendation Rate isolates those answers from passive mentions.
How to Increase Recommendation Rate?
AI engines recommend brands they trust through multiple corroborating signals: Earned media density — Tier-1 publications citing your brand as a leader, not just a participant. Across Muck Rack's three studies, 82-89% of AI-cited links trace back to earned media, the sources brands neither own nor pay for. Specificity — Content with concrete stats, named features, and differentiated positioning gives the AI something specific to recommend.
See how your brand performs in AI search
Free AI visibility audit: your AI visibility score, the publications already covering you, the gaps, and the story angles to close them.
Run Free Audit