What share of search measures, and what it does not
Share of voiceYour share of market visibility. is the brand's share of the communication in its category, and share of search is its search version: the brand's searches divided by the searches for every brand in the category. The share of voice guide reports that work presented through the IPA has argued that share of search tracks market share and may lead it, which is why it is worth tracking as a leading indicatorA measure that moves before the business outcome it is expected to predict, paired with a lagging measure that confirms the outcome later.. The guide also sets out its limits: it includes navigational searches from existing customers looking for a login or a phone number, and promotions distort it.
Branded search on its own, the volume of searches for the brand, moves with the whole category: a seasonal peak lifts every brand. The brand awareness and lift guide notes that the ratio, branded searches as a share of all category brand searches, is less affected by seasonal swings that move the whole category. That is the reason to track share rather than volume.
What share of search does not measure is memory directly. It shows interest that people expressed by searching, which mixes awareness with navigation, curiosity after a news story and reaction to offers. It is a different measure from the impression share an ad platform reports; the comparison of impression share and share of voice separates the auction measure from the market one. And it is different again from the organic visibility share that rank tracking builds from a set of queries. Share of search counts searches for brands, not positions or auctions.
Before you start
Read the share of voice guide's section on paid, earned and search share of voice, and the SEO KPIs and reporting guide, which covers collecting search data consistently enough to trust. Gather what the business already knows about its competitors, such as a recent competitor analysis, and its own market share figures if it has them, from sales data, industry sources or estimates. If it has none, the guide to estimating your market share produces a dated range on the same competitor set, with penetration kept on a separate line, ready to chart beside the trend.
Set up a sheet with one tab for term lists, one for monthly volumes, one for calculated shares and one for annotations. Decide who owns the tracking and when each month's pull happens. Treat it as a small measurement plan in the sense of the measurement plan framework: the question, the measure, the source and the review date written down before the first figure.
The steps
Define the category and the competitor set
Write down the category as the brand strategy defines it, including the alternatives customers really consider, and list the brands in it. The share of voice guide warns that a denominator of the brand's nearest rivals produces a flattering figure. Include the large brands, the challengers and any retailer or platform that customers search for as a brand in the category. Keep the set fixed; adding a brand later changes every share, so record any addition as a break.
Build each brand's term list, with misspellings and product names
For every brand, list the searches that are about that brand: the name, common misspellings and spacing variants, the name with category words people add, and well-known product or range names that people search for without the brand name. Build every brand's list the same way, so that no brand gains share because its list was more thorough. Record the date each list was built.
Exclude terms that are not about the brand
Remove searches that contain the brand's word but mean something else: a common word that shares the name, a place, a person or an unrelated product. Remove job and investor searches if they are large and unrelated to buying. Keep navigational terms such as the brand name with "login" or "contact" in a separate column rather than mixing them in, so you can see how much of the share is existing customers. Apply the same rules to every brand.
Choose one data source and keep it
The Search Console performance report, as the site's SEO guides describe it, shows only the queries a site itself appears for, so it cannot supply competitors' volumes. Use a source that covers every brand on the same basis, such as a keyword tool or public search trend data; the ad platform keyword data documented in Google Ads Help is one such source the site's guides use for search volumeThe estimated number of times a query is searched in a period, usually per month.. Different sources give different figures, so choose one, record the country, language and settings, and do not switch. If the source changes its method, log it in the measurement change log.
Pull monthly volumes for every brand
Each month, on the same day, pull the volume for every term on every list from the chosen source, using the same settings. Sum each brand's terms into a monthly total, keeping the navigational column separate. Store the raw pull as well as the totals, so that a later question about a term can be answered from the record rather than from memory.
Calculate each brand's share
Divide each brand's monthly total by the sum for all brands in the set. The shares add to the whole. Calculate it with and without navigational terms if you separated them. A ratio of figures from one source, built the same way every month, is easier to compare over time than any single volume, which is why the ratio is the measure.
Annotate promotions, launches, news and outages
In the annotations tab, record every event that could move searches for any brand in the set: your own and competitors' promotions, sales, product launches, advertising bursts, press coverage, outages and anything that sent existing customers to search for help. The awareness and lift guide asks for exactly this on branded search. Without annotations, a spike becomes an argument; with them, it becomes an explanation.
Compare year on year to remove seasonality
Many categories have seasonalityPredictable, repeating changes in demand tied to the calendar, such as months, holidays, weekdays or hours.: searches rise and fall with the weather, holidays or the school year, and some brands are more seasonal than others. Compare each month with the same month a year earlier, and read the rolling trend rather than the last month. A share that rises year on year through the seasonal dip is a stronger signal than a rise in a peak month.
Set the trend beside market share and the tracker
Put the share of search trend on the same chart as market share, where the business has it, and the awareness measures from the brand tracking survey. Look at how they moved together over time rather than at one month. Share of search and the tracker answer related questions from different directions; read together, they show whether a rise in search came with a rise in what people remember. Label the market shareThe business's sales as a fraction of all sales in a defined market, by value or by volume in one unit; a ratio of two estimates, so it is reported as a range. line with its unit and market definition, so a change in either is not mistaken for a movement. For the paid search side, the comparison of branded and non-branded search and the brand search holdout test cover what bidding on the brand's own name does and does not add.
Report the trend with its limits
Report share of search as a trend with its annotations, year on year, beside the other brand measures. State the source, the competitor set and the term rules in a footnote. Say plainly what it is: a measure of expressed interest, including existing customers, that published work argues tracks market share, not a measure of awareness itself. Avoid quoting one month's change as a result. The brand measurement hub sets out where share of search sits among the other brand measures in the same report.
The tracking sheet
The sheet below covers what the steps need. One row per brand per month in the volumes tab keeps the calculation simple.
| Tab or column | What it holds | Why it matters |
|---|---|---|
| Term lists | Every term per brand, its inclusion or exclusion, the reason and the date built | Makes the lists auditable and comparable |
| Source settings | The source, country, language and any filters, with the date of any change | Lets the series be trusted over time |
| Monthly volumes | Raw volumes per term and the monthly total per brand, navigational terms separate | The record behind every share |
| Shares | Each brand's share of the set, with and without navigational terms | The measure itself |
| Year-on-year change | Each brand's share against the same month a year earlier | Removes most seasonality |
| Annotations | Dated events for every brand in the set | Explains spikes and dips |
| Context | Market share and tracker results where available | Lets the trend be read against outcomes |
A houseplant shop tracks its share of search (illustrative)
Ploffin Houseplants tracks its share of search against five rivals
Ploffin Houseplants, an invented online houseplant shop, starts tracking share of search against 5 rivals, including a garden retailer that sells plants online. Its term list has 9 terms, including two misspellings and the name of its plant-care subscription. The brand's name shares its first letters with a common plant word, so searches for that word are excluded, which removes about 40% of the raw volume that a careless list would have counted. It chooses one public search trend source and records its settings. Over 12 months the annotations show a television gardening programme mentioning the shop in March and two sales, in May and November; each produces a spike that falls back within weeks. In the summer dip the shop's own branded volume is flat against the year before, which worried the founder, but its share of search for July and August is 14% against 11% a year earlier, because rivals' searches fell further. The brand tracker's unaided awareness, read with its interval, shows a small rise in the same period, and the report presents both as a consistent but early signal rather than a result.
Common mistakes
- A flattering competitor set. Leaving out large brands inflates every share. Define the category honestly.
- Uneven term lists. A thorough list for your brand and a thin one for rivals manufactures share.
- Counting words that are not about the brand. Exclude shared words, places and unrelated products.
- Switching data sources. Different sources give different volumes; changing breaks the series.
- Using Search Console for competitors. It shows only queries your own site appears for.
- Reading one month. Spikes come from promotions and news. Annotate them and read the year-on-year trend.
- Mixing in navigational searches unseen. Existing customers looking for a login inflate share; keep them in a separate column.
- Presenting it as awareness. Share of search is expressed interest, read beside the tracker and market share.
