# Recommendation rate

> Recommendation rate is the percentage of usable sampled answers in which an AI system endorses your brand as a choice for the buyer's need, counted once per answer.

*Source: [https://contentcucumber.com/aeo-glossary/recommendation-rate/](https://contentcucumber.com/aeo-glossary/recommendation-rate/)*

*Topic: Measuring visibility*

---

## Why it matters

An answer can name a brand as an option to avoid. A count of mentions overstates how often the brand is recommended.

## Example (illustrative, not a client result)

A brand is named in 12 of 20 answers and recommended in 5 of them, which gives a 25% recommendation rate.

## What to check

Label each usable answer as recommended, named without a judgment, or named with a caution, then divide the recommended answers by the usable answers and multiply by 100.

## What this does not tell you

We count an answer as a recommendation when the answer presents the brand as a suitable choice for the buyer's need. Because a second reviewer can label a borderline answer differently, keep the labeling rules with the report.

## Sources

- [Animalz: AEO glossary](https://animalz.co/blog/aeo-glossary) (Industry glossary)

*Evidence type: Content Cucumber reporting definition. Sources checked on 2026-11-21*

## Related terms

- [Brand mention](https://llms.contentcucumber.com/aeo-glossary/brand-mention.md)
- [AI share of voice](https://llms.contentcucumber.com/aeo-glossary/ai-share-of-voice.md)
- [Non-determinism](https://llms.contentcucumber.com/aeo-glossary/non-determinism.md)
