A New Metric for a New Discovery Channel
For decades, share of voice was the gold standard metric for measuring brand presence in media. It answered a simple question: out of all the conversations happening in your category, what percentage mention your brand? The metric worked because conversations happened in places that could be measured — press coverage, social media, advertising.
AI assistants have created a new type of conversation that traditional share of voice cannot capture. When a user asks ChatGPT "What are the best tools for brand monitoring?", a conversation happens — one that may directly influence a purchasing decision — but it leaves no trace in any media monitoring system. Share of prompt is the metric designed to measure this new channel.
Defining Share of Prompt
Share of prompt is the percentage of relevant AI-generated responses that include your brand. It is calculated by running a defined set of prompts across one or more LLMs and counting how often your brand appears in the responses.
The formula is straightforward:
Share of Prompt = (Responses mentioning your brand ÷ Total responses) × 100
For example: if you run 50 category-relevant prompts across ChatGPT and your brand appears in 18 of the responses, your share of prompt on ChatGPT is 36%.
This number is most meaningful when:
- Tracked over time to reveal trends.
- Compared to competitors running the same prompt set.
- Broken down by prompt category (awareness, consideration, decision).
- Segmented by LLM platform (ChatGPT, Gemini, Perplexity).
How to Calculate Share of Prompt Across LLMs
Calculating share of prompt manually requires three steps:
Step 1: Define your prompt set
Create a set of 20–50 prompts that represent the queries your target customers are likely to ask when researching your product category. Include a mix of broad category queries ("What are the best AI brand monitoring tools?"), use case queries ("How do I track my brand in ChatGPT?"), and comparison queries ("Compare AI visibility tools for marketing teams").
The prompt set should remain consistent across measurement periods so that changes in share of prompt reflect changes in model behavior, not changes in what you asked.
Step 2: Run the prompts and record responses
Run each prompt against each LLM you want to measure. Record the full response text, the timestamp, and the model version. For each response, note whether your brand is mentioned (yes/no) and, if yes, at what position.
Step 3: Calculate and segment
Count the number of responses where your brand appears and divide by the total number of prompts run. Calculate this separately for each LLM and for each prompt category. The resulting numbers are your share of prompt metrics — one per platform, one per category, and one overall.
For teams that want to automate this process, Promtrack runs the full calculation on a scheduled basis and surfaces the results in a dashboard with trend charts and competitor comparisons.
Share of Prompt vs. Traditional Share of Voice
Understanding the differences between these two metrics helps you use each appropriately:
| Dimension | Share of Voice | Share of Prompt |
|---|---|---|
| What it measures | Brand mentions in media and social | Brand mentions in AI responses |
| Data source | Publicly indexed content | LLM API responses |
| Update frequency | Real-time or near-real-time | Weekly or daily (scheduled runs) |
| Influenced by | PR, social activity, advertising | Content quality, authority, reviews |
| Buyer journey stage | Awareness and consideration | Consideration and decision |
| Actionability | PR and social strategy | Content, SEO, and review strategy |
The two metrics are complementary, not competing. A brand with high share of voice but low share of prompt has strong media presence but weak AI discoverability — a gap that will become increasingly costly as AI-assisted research grows. A brand with high share of prompt but low share of voice has strong AI visibility but limited broader awareness — a different strategic problem.
What Drives Share of Prompt
Understanding what influences share of prompt helps you invest in the right activities to improve it:
- Content volume and quality: Brands with more authoritative, detailed content about their product category tend to appear more frequently in AI responses. Long-form guides, comparison articles, and use case documentation are particularly effective.
- Third-party mentions: AI models weight third-party sources — review platforms, industry publications, analyst reports — more heavily than brand-owned content. A strong presence on G2, Capterra, and relevant industry publications directly improves share of prompt.
- Backlink authority: For models with web retrieval, domain authority influences which sources are retrieved and cited. High-authority backlinks improve both traditional SEO and AI visibility simultaneously.
- Recency: Models with newer training cutoffs or web retrieval favor recent content. Consistent publishing cadence maintains relevance across model updates.
Setting Share of Prompt Targets
Targets should be set relative to your competitive landscape, not against an absolute standard. A practical approach:
- Establish your current baseline and your top competitor's baseline.
- Set a 90-day target to close 50% of the gap between your current share and the category leader's share.
- Set a 12-month target to reach or exceed the category leader's current share of prompt.
These relative targets are more motivating and more achievable than arbitrary absolute targets, and they keep the focus on the competitive dynamics that actually matter.
Conclusion
Share of prompt is the defining brand metric of the AI era. It measures the thing that traditional share of voice cannot: how often your brand appears in the AI-generated responses that increasingly shape buyer decisions. Brands that start measuring and optimizing share of prompt now will build a compounding advantage in the AI discovery channel — one that will be very difficult for late movers to close.