Method6 min read
Share of search: what it predicts and when it misleads
Share of search is your brand's share of organic Google searches for all brands in its category. It tracks market share and often moves ahead of it (83% of market share on average in IPA case studies), but low search volumes and ambiguous brand names make it unreliable.
Key takeaways
- Build share of search from Google Trends topics, with your brand and every competitor in one chart for one country, so all the figures share one scale.
- Smooth weekly share of search with a 6 to 12 month rolling average before you read a trend into it.
- Put share of search beside your own market share and track the ratio over time, because it differs by category and country.
- Treat low-volume figures and one-off spikes as noise: Google says low-volume terms show as 0 and spikes on low-interest queries are not real search activity.
- Read share of search next to a monthly survey of category buyers, which shows which stage of the buying decision moved.
What is share of search?
It is a demand measure built from free search data. Les Binet, then head of effectiveness at adam&eveDDB, presented it at the IPA-led EffWorks Global conference in October 2020 as the share of organic Google search queries (“not share of paid search advertising”).
The data comes from Google Trends, which the IPA notes is free, generally goes back to 2004 and can offer weekly figures. In a 2025 Think with Google article, Binet describes the measure as “the relationship between organic search queries and market share”.
How do you calculate share of search?
Put your brand and its competitors into Google Trends as topics, in one comparison, then divide each brand’s figure by the total for every brand in the chart.
- Pick the brands. Your brand and the competitors your buyers choose between. Google’s current Explore view compares up to 8 groups of terms at once (Classic Explore compares 5), so most competitor sets fit in one chart.
- Use topics, and compare like with like. A search term returns results that include that word and none for “misspellings, spelling variations, synonyms, plural or singular versions”. Google News Initiative training calls topics “generally considered to be more reliable” because they pull in misspellings and acronyms. Compare topics with topics and terms with terms.
- Fix one country and one period, in one chart. Trends scales each chart so that “100 is the maximum search interest for the time and location selected”, so figures from separate charts can’t be added together.
- Take each brand’s share. For each week, divide the brand’s figure by the sum of every brand’s figure in the chart.
- Smooth it. Weekly figures jump about. James Hankins, who presented the IPA share of search think tank’s 2021 findings, recommends a 6 to 12 month rolling average.
- Set it beside your market share. The ratio between the two differs by category and country (see below), so your own category’s ratio is the baseline to watch.
How well does share of search predict market share?
Closely on average, and often in advance, but the evidence is correlation and the ratio differs by category.
Binet’s 2020 work covered three categories: automotive, energy (gas and electricity) and mobile phone handsets. Share of search correlated with market share in all three, and moved first: when it rose, market share tended to rise after it, and when it fell, market share tended to fall. “The lead time for Share of Search on Share of Market can be substantial, up to a year for cars.”
He called the gap between the two Extra Share of Search (ESOS), “a particularly solid indicator of market share movements”. In 2025 he wrote that “movements in share of search often precede movements in market share”.
In 2021 the IPA think tank tested the idea on 30 case studies representing 12 categories across seven countries:
- 83% on average. Share of search “appears to represent 83% of a brand’s SoM on average”.
- The ratio varies “by category and country”. Check your own category’s ratio rather than assuming 83%.
- The two move together. “When one changes the other changes.”
- No proof of cause. The findings “refer to correlations and not causal relationships”.
- Going dark costs share of search. Stopping communication “seems to result in a consistent pattern of decline”, related to how often buyers purchase in the category.
The think tank found that “when a brand spends more than its market share and increases its share of voice, its share of search also increases”. Binet found the reverse: if share of voice falls below a brand’s equilibrium level, “the Share of Search tends to fall over the next two years”. That makes it evidence for the budget case in investing in your brand.
When does share of search mislead?
In six cases: five from Google’s Trends help pages or Binet’s caveats, one from our reading of the IPA results.
Low volume. “Trends only shows data for popular terms, so search terms with low volume appear as ‘0’.” Google also adds statistical noise to protect privacy, “most noticeable on queries with low or no search interest”, and says the resulting one-off spikes “should not be interpreted as real search activity”. A small brand’s share can be mostly noise: smooth heavily, or don’t use the measure.
Ambiguous names. A search term matches any query that includes the word. If your brand name is also an everyday word or a place name, a term counts all those searches as yours. Use the topic; if Trends has none for your brand, don’t trust the number.
Searches Trends leaves out. Trends removes repeated searches from the same person over a short period, “queries with apostrophes and other special characters”, and “internal searches made by AI Mode and AI Overviews”. If your brand name has an apostrophe, prefer the topic.
Searches that aren’t demand. Binet warned that “not all search activity is positive”. A product recall or a wave of complaints, for example, raises your searches without raising your sales. Conversion is also “affected by other factors, particularly price”, so buyers can search for you and still buy a cheaper rival.
Categories where buyers don’t search (our reasoning). The IPA found the ratio to market share “varies by category and country”, so test the measure against your own market share history before you trust it. Where buyers rarely search before they buy, or hand the task to an assistant (see how AI shopping agents choose brands), we would expect the search record to cover less of the decision.
Reading it as a poll. “Google Trends is not a scientific poll and shouldn’t be confused with polling data”. Google says the data “should always be considered as one data point among others before drawing conclusions”.
What should you read next to share of search?
A survey of the people who buy in your category. Share of search shows that interest moved; the survey shows which part of the decision moved with it.
Ask about the same brands you put in Trends:
- Unprompted awareness: how many category buyers name you without a list.
- Prompted awareness: how many recognise you when shown your name.
- Consideration: how many would think about buying from you next time.
- Preference: how many choose you over your closest rivals.
If share of search rises while consideration and preference stay flat, the extra searches are not turning into choice; brand awareness is up and sales are flat shows how to find where buyers drop out. Wink, the brand tracking platform, runs this survey every month with real category buyers, comparing your brand with up to five competitors.
Build the chart this week: your brand and its competitors as topics, one country, the longest period the data allows, smoothed over 12 months. Write your latest market share figure beside it, so next quarter’s movement has a ratio to be measured against.
Common questions
Does share of search include paid search?
No. Les Binet defined it as the share of organic Google search queries, not share of paid search advertising.
Where does share of search data come from?
Google Trends. The IPA notes the data is free, generally goes back to 2004 and can offer weekly figures.
What is extra share of search (ESOS)?
The gap between a brand's share of search and its share of market. Binet found it is a particularly solid indicator of market share movements.
Is share of search the same as share of voice?
No. Share of voice is a brand's share of advertising in its category; share of search is its share of brand searches. The IPA think tank found that when a brand spends more than its market share and increases its share of voice, its share of search also increases.
How far ahead of market share does share of search move?
Binet found the lead time can be substantial, up to a year for cars. Check the lag in your own category by comparing past share of search with past market share.
Sources
- Binet presents fast, cheap, predictive Share of Search metric (IPA, 14 October 2020)
- New findings from the cross-industry IPA Share of Search think tank data (IPA, 12 October 2021)
- FAQ about Google Trends data (Trends Help)
- Compare Trends search terms (Trends Help)
- Basics of Google Trends (Google News Initiative)
- Les Binet: How econometrics is shaping the future of measurement (Think with Google UK, March 2025)