How to Calculate
Using the definition of Share of Voice, we can calculate SOV using the following formula:
(Number of mentions of your brand / total number of brand mentions) X 100 = SOV %
This simple calculation applies to the spend or visibility basis, for a product or service in a defined market. The result is a percentage (%) metric that can be used as a digital marketing Key Performance Indicator (KPI).
The trick to SOV calculation lies in the data sourcing and the definition of spend or visibility and to the defined market or channel. The practical calculation method depends on the use case or the model used.
Use Cases/Models
The intuitive simplicity of SOV calculation can be deceptive, and the reality of data sourcing, data capture and data processing in each channel require care. The validity of each channel SOV method should be assessed as to the data availability and the target of analysis.
In principle the variations of SOV calculation apply to how they are used and their use case.
Share of Search
Share of Search (SoS, or Share of Search Engines) is touted as the pragmatic answer to SOV calculation by virtue of the rise and dominance of search engines in the user journey.
(Share of your brand in Google / total sector in Google) X 100 = SoS %
This applies equally to Bing and other search engines, but we’ll focus on the ubiquity of Google for the sake of illustration. Rankings data is collected for your brand and all your competitors – every website that ranks in the results.
Simple in theory, more complicated in practice when isolating and measuring the parameters.
Google search result rankings (/positions) can be measured for a defined set of target keywords which together are representative of the market – this can be the whole market as best as can be defined by a large data set. The larger the set the more accurate a reflection of the entire sector it may be, but it will encroach upon other sectors and products in doing so. Smaller sets of target terms have an advantage in being more specific SOV to targets, though become unreliable for longer term analysis of excess SOV.
For a defined set of target terms, SoS can be calculated for total search or broken down to the individual search channels of Organic (/natural) and Paid Search:
(Share of your brand in Google Organic / total sector in Google Organic) X 100 = Organic SoS %
and
(Share of your brand in Google Paid / total sector in Google Paid) X 100 = Paid SoS %
Share of search engine result rankings (/positions) requires consideration of further factors in accurate data collection:
- multiple positions (not just top position as is the focus of SEO).
- other key search results (inc. Google Knowledge Graph/GKG and featured snippets).
- the click through rate (CTR) on search result positions.
- the relative search volume for keyword targets.
The determination and inclusion of each of these other factors in search results is important to understand the scope and interpretation of SOV.
Google Knowledge Graph/GKG is a high visibility return with high impact on CTR and should be accounted for its impact on brand visibility.
Some variations of SoS calculation use Google Trends data as a replacement when ranking positions data cannot be determined or captured. A relative share is used as a proxy. However, the consistency and completeness of Google Trends data, compared to directly sourced SERPs undermine its validity.
NB: SoS is not to be confused with a share of the search engine market, where for instance Google might be used for 80% of searches in one market (which would be a share of market measure and not a Share of Search metric of brand voice to potential users of search engines).
Share of SEO
SEO is normally considered to apply to Organic (/natural) search results only. Taking only organic ranking positions into account would lead to a Share of SEO.
(Share of your brand in Google Organic / total sector in Google Organic) X 100 = Organic SoS %
This would be in parallel to the paid search channel and an equivalent and complementary calculation of share of paid search.
Google has specific metrics for of impression share of paid, Google Impression Share.
Bing similarly has metrics for paid search – Microsoft Impression Share
Within both channels the difference between branded and unbranded (/generic) search terms can be further considered. The benefit of such further breakdown of the data providing insights into the structural and behavioural aspects of search optimisation, whilst again losing the benefit of long term eSOV through greater granularity.
Share of eCom
eCom sites such as Amazon are also online search origination points for user search behaviour. Calculation of SOV within eCom sites is possible through data measurement of the same dimensions as measured within search engines.
(Share of your brand in Amazon / total sector in Amazon) X 100 = eCom SOV %
Search result rankings (/positions) in Amazon are used in the basis of this calculation, again:
- multiple positions
- organic and paid ranking positions
- the click through rate
Again distinction can be made between, or at least data captured for both, branded and unbranded (/generic) search terms to measure an accurate and actionable scope for eCom SOV.
Share of Social Media
Taking Social Media as another channel, brand mentions – this time as impressions rather than relative positions in results – can be measured to form a percentage of all mentions in a market.
The channels including Facebook, Instagram, YouTube, Twitter, LinkedIn will only be as good as the source data that can be collected.
Data difficulties are introduced given the number of social channels. Also impression measurement can be complicated by ‘rant factor’, in which it is more difficult to determine a positive from a negative mention – leading to a more qualitative measurement of share in calculation of SOV.
Share of Target
Within an identified and measurable channel, particularly relevant to online search within Google or Amazon, a further subset of terms or ads, can be selected for data capture and SOV measurement. This method negates the need to deal with the difficulty of measuring ‘an entire market’ and how this might be defined in practical terms.
For a specific set of target keyword search terms, SOV can be calculated and used by marketers as a KPI on a campaign basis (normally over a set period of time instead of the long term).
(Share of your brand for target terms / total sector for target terms) X 100 = SoT %
Automation of SOV
The automation of paid share data is available in Google and Bing according to their definition of the metrics available in their bidding interfaces. This can be fed into and visualised in Google Data Studio but is restricted to paid only SOV.
Automation of organic share is provided by providers now (including BrightEdge, but not paid) and both organic and paid SoS is available from others (including Cressive DX).
All automation of SOV calculation relies on the methods and reliability of data sourcing and capture for the channel, sector or market in question. Followed by the processing and presentation of SOV data into systems that delivers the metric to marketing teams in a timely digestible way as part of their business as usual analysis.
The old adage of ‘rubbish in, rubbish out’ is worth reciting when evaluating competing methods of SOV calculation automation.
At a wider level of marketing advertising, Nielsen and Kantar offer SOV metrics, the automated level of this depends on the advertising sources.