How it works
How Share of Conversation works.
Shoppers are asking AI what to buy. Share of Conversation puts a certified set of real shopper questions to Gemini, OpenAI and Perplexity, recognises the brands put forward, and measures each brand’s share of the conversation against its competitors. The wider AI Commerce Benchmark can add other AI environments, measured separately where included.
The shift in commerce
From search results to AI answers.
Instead of scrolling through pages of results, shoppers increasingly receive a short, curated answer — from direct AI assistants, from AI inside search, and from retailers’ own AI. Each can put forward a different set of brands.
Search results
Shoppers scroll through ranked links and product listings, and choose from what they see.
and, increasingly
AI answers
Illustrative
?
Shopper question
Shoppers ask a natural question…
AI answer
…and receive a curated answer…
Your brand
Brands put forward
…that names specific brands.
Where your brand appears in those answers now matters.
Our process
Five steps from shopper question to competitive evidence.
01
We build a certified set of questions that reflect how shoppers actually ask about your category — from broad needs to specific missions.
02
Each question is put to the Direct AI providers Gemini, OpenAI and Perplexity in a consistent way. Where a study includes them, AI Search and Retailer AI are measured separately.
Gemini
OpenAI
Perplexity
03
We identify every brand referenced in each answer and match it to the category’s competitive field. Where a reference can’t be recognised with confidence, it is handled conservatively.
Recognised
Unrecognised
04
Share of Conversation measures a brand’s share of recognised eligible brand references across a certified universe of shopper questions. Every brand is measured against the same category evidence base — one clear measure, read against the whole category.
Share of Conversation
Your brand
Whole measured field
05
We show where your brand appears, how the field changes by provider and by shopper need, and where observable differences are worth investigating.
Your brand
The core measure
One measure. Different ways to read the field.
One core KPI, read by provider and by shopper need. No composite scores, no indices.
Core KPI
A brand’s share of recognised eligible brand references across a certified universe of shopper questions.
Your brand
Whole measured field
By provider
Illustrative
Provider view
Gemini
OpenAI
Perplexity
The same Share of Conversation, read separately for Gemini, OpenAI and Perplexity — showing how the competitive field changes from one AI to the next.
Direct AI
By shopper need
Illustrative
Shopper-need view
Question group
Question group
Question group
Share of Conversation read across groups of related shopper questions — showing where the field shifts as shopper needs change.
Where measured
Real evidence · Sun Protection & SPF UK
One benchmark, from question to measure.
Step 01
160
certified shopper questions
Step 02
480
Direct AI answers
Step 03
2,087
counted brand references
Step 04
172
brands in the measured field
Share of Conversation · Direct AI
La Roche-Posay
Sun Protection & SPF UK
In our Sun Protection & SPF UK benchmark, La Roche-Posay held 15.24% Share of Conversation — measured across 160 certified shopper questions and 480 Direct AI answers, from 2,087 counted brand references spanning 172 brands in the measured field.
Figures are specific to this category and not comparable across categories.
Reliability
Evidence you can stand behind.
The disciplines behind every benchmark.
Certified
Consistent
Conservative
Observation period
The framework
From AI evidence to category insight.
How Share of Conversation turns AI answers into evidence a commercial team can use.
AI evidence
Certified shopper questions, Direct AI answers and recognised brand references — measured the same way, every time.
Competitive intelligence
The core KPI, measured across Direct AI and read by provider and by shopper need.
What to investigate
Where your brand appears, how the competitive field changes, and which differences are worth investigating.
AI environments work differently, and some operate as closed ecosystems with limited transparency. Shopper demand, AI Search, Retailer AI and the product / digital shelf are each measured on their own terms, where evidence is available — and never blended into Share of Conversation.