Share of search
Also: SoS, share of search, SOS, part de recherche, search share
The proportion of searches in your category that mention your brand, a leading indicator of market share.
What it is
Share of search is the percentage of total search volume in a defined category that is captured by branded queries for your company. If people run 100,000 category searches in a month and 12,000 of them include your brand name, your share of search is 12%.
The metric is usually built from search engine query data (via keyword tools or platform APIs). You define a category by grouping the generic queries that describe it (for example "running shoes", "trail shoes", "marathon trainers") and the branded queries for every competitor, then divide your brand volume by the total.
Why it matters
Share of search tends to move ahead of market share, which makes it a leading indicator rather than a lagging one. Rising consumer intent shows up in search behaviour before it shows up in revenue or reported share.
- It is available weekly or monthly, far faster than syndicated market share panels.
- It is relatively cheap to track and covers competitors you do not have sales data for.
- It captures demand and attention, not just conversion, so it helps separate a brand problem from a distribution or pricing problem.
How it is used in practice
- Forecasting: correlate historical share of search against market share to estimate future share.
- Campaign measurement: watch for lifts in branded search after advertising or PR, a proxy for mental availability.
- Competitive tracking: benchmark your trend line against rivals to detect share shifts early.
- Early warning: a sustained decline often precedes a revenue decline by one or more quarters.
Caveats: category definition drives the result, so keep the query set stable. Generic and branded queries must be classified consistently, and voice, app, and marketplace search sit outside most datasets.
Worked example
A category has 500,000 monthly searches. Branded volumes:
- Brand A: 150,000 (30%)
- Brand B: 120,000 (24%)
- Brand C: 80,000 (16%)
- Unbranded / others: 150,000 (30%)
Brand A holds a 30% share of search. If its current market share is 26%, the 4 point gap suggests demand is running ahead of sales, pointing to a distribution or availability constraint worth investigating.
Frequently asked questions
How is share of search calculated?
Divide the monthly search volume for your branded queries by the total search volume of the category, branded and generic combined. If a category generates 100,000 searches a month and 12,000 of them include your brand name, your share of search is 12%. The category is defined by grouping the generic queries that describe it plus the branded queries of every competitor.
What is the difference between share of search and market share?
Market share measures sales already realised; share of search measures search intent, and it tends to move first. That timing difference is the point: a sustained drop in share of search often precedes a revenue decline by one or more quarters, so it works as a leading indicator rather than a lagging one.
Who uses share of search, and for what decisions?
Marketing, data and finance leaders all use it. Marketing tracks lifts in branded search after a campaign or PR push as a proxy for mental availability, data teams correlate the series with market share to build forecasts, and finance uses the trend line as an early warning on revenue. It also covers competitors for whom you have no sales data.
What does it mean when share of search is higher than market share?
Demand is running ahead of sales, which usually points to a distribution, availability or pricing constraint rather than a brand problem. Example: a brand with 30% share of search and 26% market share has a 4 point gap worth investigating on the sell-side, not on the awareness side.
What are the main limitations of the metric?
The category definition drives the result, so the query set has to stay stable from one period to the next, and generic versus branded queries must be classified the same way every time. Voice search, in-app search and marketplace search sit outside most keyword datasets, so a share of search figure describes search engine behaviour only.