What Is Share of Search?
Share of search is a brand's percentage of the total branded search volume within its competitive set — your branded queries divided by the category's branded queries combined. Popularized by Les Binet's research presented at IPA EffWorks in 2020, it earned its status with a striking property: share of search predicts market share movements roughly 6-12 months ahead in categories from automotive to mobile handsets, making it the cheapest leading indicator in marketing analytics.
Why it predicts market share
Buyers search for brands they are beginning to consider; aggregated, that intent precedes purchases by months. Binet's work (building on the Ehrenberg-Bass mental-availability tradition) showed the correlation across long time series, and the metric's inputs are free — Google Trends and Keyword Planner — which is why it spread from econometricians to every competitive dashboard.
Extending the construct to AI prompts
The AI-era analog replaces queries with prompts and adds a second layer:
| Layer | Classic version | AI extension |
|---|---|---|
| Demand side | Branded search volume share | Branded prompt share (what users ask assistants) |
| Supply side | — (SERPs don't volunteer brands) | Share of recommendation in unbranded answers |
That supply side is genuinely new: an assistant answering "best CRM for a nonprofit" volunteers names, shaping consideration before any branded demand exists. Prompt market share captures this recommendation layer; classic share of search captures the demand layer. Watching both reveals sequence — recommendation share tends to move first, then branded demand, then revenue.
How to operationalize it
- Fix the competitive set (5-10 brands) and pull monthly branded volumes.
- Compute and trend the share; annotate campaigns and launches.
- Run the parallel AI measurement: sample category prompts across engines and score recommendations via competitor analysis.
- Investigate divergence — rising AI recommendation share with flat search share often means visibility gains haven't yet compounded into demand.
Example: a challenger fintech sees its share of search climb from 9% to 14% over three quarters while AI recommendation share had already doubled two quarters earlier — the leading indicator now has its own leading indicator.
Frequently asked questions
- How is share of search calculated?
- Divide your brand's search volume by the total branded search volume of the competitive set: your volume / (yours + all tracked competitors'). Google Trends comparisons or Keyword Planner exports across the brand list are the standard sources; measure monthly and trend it.
- Does the share-of-search logic really transfer to AI prompts?
- The mechanism transfers: both measure how often buyers reach for your name unprompted, which reflects mental availability. The prompt version adds a twist — assistants also volunteer brands in unbranded answers, so AI-era share combines what users ask for with what engines recommend.
Keep exploring
See how AI engines talk about your brand — track mentions across ChatGPT, Perplexity, Claude, Gemini and 5 more. Start with Menra