Measure AI share of voice against your competitors
Your board asks whether AI assistants recommend you or your rivals. Measure share of voice and citations against named competitors on six AI engines, brand by brand, and pull the data into your own stack.
Custom plans: volume pricing, dedicated support and an SLA.
fieldnote.app4,571 creditsFN
Competitors
Brand: Fieldnote · 40 prompts × 6 engines
Illustrative data
Share of mentions
31%
+4 pts vs Q2
Share of citations
22%
+2 pts vs Q2
Prompts led by a rival
14
of 40
Share of mentions by brand
Pipewise38%
Fieldnote31%
Hubstack19%
Loopdesk12%
What is AI share of voice?
AI share of voice is the share of brand mentions, or citations, your brand wins across a fixed set of prompts and AI engines. Run the same prompts on ChatGPT, Claude, Gemini, Perplexity, Google AI Overviews and AI Mode, count how often each brand is named, and divide by all brand mentions in those runs. Repeat each quarter with the same prompt set so the trend compares like with like. Enterprise plans, access and governance
Why AI share of voice is a blind spot
Search share and brand trackers stop where the AI answer starts. Leadership is asking about a channel nobody owns yet.
No measured competitive baseline
Anecdotes from someone's ChatGPT session aren't a metric. You need the same prompts, run on the same engines, stored over time, for you and named rivals.
Many brands, many markets
Each brand and product line has its own competitors and prompts. A single dashboard view hides the brand that is slipping.
Data trapped in a UI
Insights teams want the numbers in their warehouse, BI tool or AI assistant, not in screenshots.
Access and governance
Regional teams and agencies need access to their brand only, with clear roles and controlled spend.
The prompts that decide your category
Competitive intelligence starts with the questions buyers ask when they are choosing between you and a rival.
A competitive prompt set per brand
Build prompts for category leadership, head-to-head comparisons, alternatives and enterprise requirements. Each run records whether you and each competitor were named, at what position, with what sentiment, and which URLs were cited.
Head-to-head prompts show where a rival is preferred
Alternatives prompts show who wins switching buyers
Add each brand as a domain, name its direct competitors and save its prompt set. AI Prompt Research can suggest prompts for a chosen country and language.
Run the prompts on ChatGPT, Claude, Gemini, Perplexity, Google AI Overviews and AI Mode. Every run is stored, so share of mentions and citations can be compared across quarters.
Citations shows which domains and URLs each engine cites for each prompt, including competitors' pages and third-party reviews. That is where share of voice is won.
Progress Reports compare two points in time, or your brand against a competitor, as a PDF with an executive summary. The Action Plan turns gaps into prioritised work for content and PR teams.
Progress report · Q3 vs Q2Illustrative
Executive summaryIncluded
Brand vs Pipewise comparisonIncluded
Work completed this quarterIncluded
The features enterprise teams use
Measured data for decisions, AI-written analysis for ideas, and open access for your own stack.
Measured competitor visibility
Brand visibility records every competitor named in each answer, so share of mentions, position and sentiment come from stored runs, not estimates.
An AI-generated comparison of each competitor's site against yours: content gaps, technical advantages, keyword opportunities and recommendations. Use it for hypotheses; confirm with the measured data above.
Enterprise security page with SOC 2 detailContent gap
Comparison pages for 5 rivalsContent gap
Faster product page loadTechnical
API and MCP access
Pull visibility, audit, keyword, SERP, GA4 and Search Console data into your own tools through the REST API, or query it from Claude through the MCP connector.
Visits and conversions from AI assistants, from the tracking tag and GA4.
44%
of prompt–engine pairs switched between naming a brand and leaving it out at least once. Board metrics need repeated runs.991 successful runs of 45 prompt–engine pairs tracked in AI Rank Lab, 13 February to 30 October 2026. Pairs with five or more runs only. Aggregated and anonymised.
What to keep out of the board pack
Measured numbers and AI-written analysis both have a place. Label which is which.
Competitor Analysis is AI-written
Its scores are Gemini assessments of competitor sites, useful for hypotheses. Share of voice for the board should come from measured brand visibility runs.
Prompts run when you start them
There is no automatic daily monitoring. Run the prompt set before each review; prompts opted in to the Action Plan citation tracker are re-checked weekly.
Assistants don't emulate a location
Runs on ChatGPT, Claude, Gemini and Perplexity don't pretend to be in a country or town, so write the market into the prompt. Google AI Overview and AI Mode checks take a country and language per query.
One run is not a sample
In our data, 44% of tracked prompt-engine pairs flipped between naming a brand and not at least once. Report share of voice across several runs per prompt.
It is measuring how AI assistants and AI search answers describe your category: which brands they recommend, in what order, with what sentiment, and which sources they cite, for you and your named competitors.
Which enterprise AEO platforms measure AI search performance?
Look for one that runs your prompts on several engines, stores every run, records competitors and cited URLs, and exposes the data by API. AI Rank Lab does this for ChatGPT, Claude, Gemini, Perplexity, Google AI Overviews and AI Mode. It is our product, so test it against your own prompts.
How is AI share of voice calculated?
Over a fixed prompt set, count how often each brand is named (or cited) across runs and engines, and divide by all brand mentions (or citations) in the same runs. Keep the prompt set stable between periods so the trend is like for like.
Are prompts run automatically every day?
No. You run prompts when you choose and every run is stored. Prompts opted in to the Action Plan citation tracker are re-checked weekly.
Can different teams see only their brand?
Yes. Share a domain with a person at the employee, client or viewer role, choose which features they can use, and allocate credits to that domain.
Do brand visibility runs reflect a specific country?
Google AI Overview and AI Mode checks take a country and language per query. Runs on ChatGPT, Claude, Gemini and Perplexity do not emulate a location, so write market-specific prompts (for example, naming the country) where it matters.
Can we pull the data into our BI tools?
Yes. Visibility, audit, keyword, SERP, GA4 and Search Console data are available through the REST API, and from Claude through the MCP connector.
What does enterprise pricing look like?
The Enterprise plan is $499 a month for 150,000 credits. Custom plans add volume pricing, dedicated support and an SLA. Book a demo to scope it.
