Which brand does the agent pick — and why?

---
title: Which brand does the agent pick — and why?
canonical: https://chosenbyagent.com/questions/which-brand-does-the-agent-pick
updated: 2026-10-22
audience: brands
offers:
  - B3
evidence:
  - larger-store-pick
  - rating-sensitivity
  - model-update-reshuffle
  - api-consumer-overlap
---

# Which brand does the agent pick — and why?

Short answer: Mostly the larger, better-rated option, and the winner can change when the model does. Being mentioned is not the same as being chosen, so we measure the choice.

## What the research shows

- **90–94%** — In unlabelled two-store comparisons, assistants picked the larger store 90–94% of the time. [ALM Corp, 2026, vendor](https://almcorp.com/news/ai-shopping-answers-chatgpt-google-lean-toward-big-retailers/)
- **30–80 pp** — A product rating shifted model choice by 30–80 percentage points in 14 of 17 models, against 5 points for people. [ABxLab, arXiv 2509.25609, 2025, independent](https://arxiv.org/html/2509.25609)
- Moving Gemini 2.5 Flash from preview to general availability redistributed choice shares and changed the most-chosen product in two categories. [ACES, arXiv 2508.02630, 2025, independent](https://arxiv.org/pdf/2508.02630v3)
- **15–32%** — Brands named in API answers overlap with what consumer assistant apps show on only 15–32% of brands. [Surfer, 2026, vendor](https://surferseo.com/blog/llm-scraped-ai-answers-vs-api-results)

## Choice test

B3 · from €3,000 (per category) · indicative · scoped per engagement

Who wins your category and why — mentioned versus chosen, sensitivity to price, reviews and returns.

- 3–5 scenarios, 3–4 models, several personas
- Mentioned versus chosen
- Sensitivity to price, reviews and returns policy
- Confidence intervals; reported as an index, not market share

## What you get

Three to five scenarios in your category run on three to four models with several buyer personas. We report who is mentioned, who is chosen, and what moved the choice.

**Mentioned vs chosen, one category**

| Brand           | Mentioned | Chosen | Main reason                  |
| --------------- | --------- | ------ | ---------------------------- |
| Your brand      | 62%       | 21%    | higher price than the leader |
| Category leader | 88%       | 54%    | rating and review count      |
| Challenger      | 41%       | 18%    | free returns                 |

_Illustrative values — your report shows yours._

## What this does not claim

- Results are an index with confidence intervals, not market share.
- If an agent buys, this is who it picks — we do not estimate how often agents buy.
- About 325 runs per cell give ±10 points; tighter intervals cost proportionally more.

## Request this instrument

We reply with a scope, a fixed price and a start date.

Agents: POST /api/engagements (application/json) with {"offer_id": "B3", "on_behalf_of": {"email": "…"}, "agent": {"name": "…"}}. The person confirms by email before we see the request. Details: /agents.md

### Related questions

- [What changes if I change price, returns or copy?](/questions/what-if-i-change-the-offer.md)
- [Can a shopping agent actually get through my site?](/questions/can-an-agent-get-through.md)
- [What changed after the last model update?](/questions/what-changed-after-the-update.md)