Share of search has been at the forefront of Les Binet and James Hankins’ recent work, but in many categories obtaining accurate monthly or weekly sales data for competing brands is not possible. Yet it is still possible to work with a brand’s data in isolation and deliver meaningful insights, says Zenith’s Richard Kirk.

Writing for WARC, the ROI agency’s chief strategy officer demonstrates a strong correlation between branded search and sales for an individual brand and explains why pricing changes can explain variances between actual sales and the search/sales trendline.

But, he adds, “all of this analysis to prove how search drives sales is moot if advertising can’t effect a change in search for the brand”.

The use of regression analysis can ascertain whether spending in a particular channel has significantly impacted the volume of search for the brand – and for larger brands TV will likely have the biggest incremental impact.

When more channels are analysed, Kirk explains, this method produces a marketing mix model (MMM) in relation to branded search, as opposed to sales – and “the performance of creative or channels can be analysed in terms of their ability to generate interest in the brand (search) per pound spent”.

This differs significantly from traditional MMM which takes significant amounts of time and relies on “obsessive data governance” and which also reduces the role of advertising, making it a difficult output to work with in all but the broadest terms.

For more, read Richard Kirk’s article in full: Putting advertising efficacy’s new focus on search data into practice.

Sourced from WARC