What’s Your Share of Search?

By Giles Martin, EVP, Strategy & Insights
Oxford Road

Les Binet, one of the “Godfathers of Advertising Effectiveness”, often creates a lot of attention in the marketing community when he publishes new research. We believe this attention is well-deserved: Binet’s work with Peter Field has arguably changed the industry forever, ushering in a new era of accountability and clarity for marketers, and demonstrating what will drive growth for their organizations.

Last month, Les created waves with a stimulating, promising, and challenging video about Google search data.  Broadly speaking, he claims that “Share of Search” data can be used to predict (or, at least, may be useful in predicting) market share data for brands. For reference, share of search is defined as the search volume for Brand X divided by the total number of searches in the category.  

Binet and his team looked at three different categories to provide some range in the analysis: automotive, energy, and mobile phones. Here’s an example of how share of search data correlates with market share in the automotive sector. You can see immediately it’s a pretty strong and clear correlation.

What’s essential about Binet’s latest contention is not the correlation per se, however. Rather, it’s the relationship between the time series of these data sets. Below is an example showing the trend across ten years of data for share of search and market share for LG in the smartphone market. Binet suggests that the share of searches is a valuable leading indicator for future trends in market share, which is certainly implied by the chart.

He goes on to claim that the increase in share of search, across the previous nine months, can predict an increase in market share in the current quarter. The implication, then, is that changes in share of search can be used as a type of early warning indicator for brands.

Another important piece for marketers to understand is the relationship between advertising and the share of searches. The chart below shows a strikingly strong relationship between the two data series (or specifically, changes in increase in SOV correlating with changes in share of search volume.) Said another way: if you spend enough in offline marketing to meaningfully increase your share of voice, you’ll drive an increase in the share of searches in your category. It makes a lot of sense if you think about it.

This type of analysis will be familiar to some. Binet & Field popularized for many the view of the relationship between “excess” share of voice and market share gain, providing many advertisers with valuable guidance on budget setting to achieve specific financial and business goals.  This latest work connects the key dots between these two variables: how broad-based (generally offline) advertising investment creates an impact on people’s neurology, which in turn manifests in more curiosity, awareness, and consideration for the brand. This increase in mental territory and activity for the brand, visible through heightened search activity, finally turns into revenue and market share. 

Despite the promise and interest of this approach, it’s still in an early phase. We advise a degree of skepticism, or perhaps cautious optimism, about the promise of this data. It remains the first work of its kind (ie. using share of search), and while three categories are better than one, it’s still only three categories. A good next step will be widening the category set and making some of this data more publicly available. 

What should advertisers and clients be doing with this information? Well, firstly and simply, they should be aware of this work and its importance. Secondly, especially for D2C clients, they should view this work as a reminder that being a “data-driven” marketer is more than looking at a portfolio of channel performance. It’s also about looking at data sets and marketing research that provide pointers – increasingly valuable and robust pointers – about how marketing works and how it can drive growth in a broader sense for their corporations.