AI-driven commerce discovery
How Share of Conversation Works
Consumers are no longer just searching — they are asking AI what to buy. Share of Conversation measures how often your brand appears in AI-generated recommendations and how it compares to competitors.
The shift in commerce
From Search to AI Recommendations
Instead of browsing pages of results, consumers now receive curated recommendations from platforms like ChatGPT, Perplexity, Amazon Rufus, and Google.
Traditional Search
Consumers scroll through pages of ranked links and product listings.
Our Process
Five Steps to Measuring AI Visibility
We build structured prompt sets based on real consumer intent — the exact questions shoppers are asking AI platforms today.
Each prompt is run across multiple AI and search platforms. We capture which brands are recommended, how often they appear, and how prominently they are positioned — creating a consistent view of AI-driven visibility.
SoC represents the percentage of AI recommendations your brand captures within a category. It's calculated using frequency of mentions, breadth of prompt coverage, and relative positioning vs competitors.
Your performance is always contextualised — who is leading, where you are winning, and where you are being missed. This highlights real competitive gaps in AI discovery.
We connect visibility gaps to commercial impact: missed high-intent prompts, weak positioning vs competitors, and opportunities to improve product content and signals.
The Metric
How We Calculate Share of Conversation
All metrics are derived from the same underlying dataset of prompt runs and AI-generated recommendations.
SoC %
Share of Conversation
The percentage of AI-generated recommendations your brand captures within a category.
SoC = (Brand mentions ÷ Total mentions) × 100
ADS
AI Discovery Score
A weighted score reflecting how strongly your brand performs across AI discovery signals — visibility, coverage, position, and frequency.
Weighted: SoC + Coverage + Position + Frequency
PC%
Prompt Coverage
How many relevant shopper prompts your brand appears in out of the total prompts analysed.
PC = (Prompts with brand ÷ Total prompts) × 100
AP
Average Position
Where your brand typically appears in AI recommendations — the mean ranking position across all prompts.
AP = Mean rank across all appearances
High-Value AI Discovery Gaps
High-Value Missed Opportunities
High-value prompts where competitors are recommended but your brand is missing or underrepresented.
Competitor SoC > 25%
Your Brand SoC < 15%
41%
Example
Pampers appears in 41 out of 100 AI-generated recommendations → SoC = 41%. An AI Discovery Score of 82% indicates strong visibility across prompts, consistent high positioning, and frequent inclusion relative to competitors.
EXAMPLE
"Some AI recommendation environments operate as closed ecosystems with limited transparency. Our methodology adapts measurement approaches by platform.”
Reliability
Data Confidence
Every dataset is assigned a confidence level based on the number of prompts analysed, platform runs, and consistency across sources.
High
50+ runs, 10+ prompts, multiple platforms
Medium
15+ runs, 5+ prompts
Low
Below threshold or demo data
Important:
AI platforms are dynamic and constantly evolving. Share of Conversation reflects observed patterns at a point in time, not fixed rankings.
The Impact
Why It Matters
Brands that appear more often in AI recommendations gain a decisive advantage in the new discovery landscape.