Refining search and social data to be more useful

by Nigel Hollis

Marketers today need fast feedback on whether their campaigns are likely to succeed or not. While automated pre-testing provides the opportunity to anticipate how well a campaign will perform, many will still want to check whether the in-market performance is on track for success and properly analyzed search and social data can help provide that feedback.

Everyone knows that the digital world offers a ton of free data on what people are doing and saying online. The only problem is that data alone is not useful. Data is like oil; it needs to be refined to be useful. So to make the right decisions about your brands you need to be sure you are answering the right questions with the right data at the right time. This is why Kantar Millward Brown has invested time and analytic resources to make sure that we can extract useful information from both search and social data as part of its Holistic Brand Guidance Systems.

In order to make search and social useful we have to take a number of basic considerations into account. First off we need to recognize that the two data sets are telling us different things:

•     Social Campaign Efficiency indicates how successful a campaign is at driving brand fame
•     Search Campaign Efficiency indicates how successful a campaign is at generating relevant interest

Then, we all know social and search data can be messy and confusing and it is difficult to determine what it means for brands but the team at Kantar Millward Brown have managed to use a sophisticated modeling approach to extract actionable insight for brand owners (and in an affordable and scalable way). The modeling allows us to isolate the response to the current campaign from other systemic influences on behavior (more on how we use dynamic linear modeling to do so here). Once that is done, we need to take the nature of the brand and its history into account.

Social Campaign Efficiency needs to take into account how much people are likely to talk about the brand anyway. For Search Campaign Efficiency we need to take into account how likely advertising is to spark searches in the advertised brand’s category. People are far more likely to search for brands that operate online, like airlines, banks and online retailers, than those where interaction is largely offline.

This analysis allows us to quickly report the impact of a new campaign and put it into the context of the brand’s previous campaigns as well as a general brandscape. As a result decisions can be taken on whether a campaign is working as effectively as it could and what to do about it. Reassuringly we find that there is a good relationship between the Campaign Efficiency metrics and other survey metrics. High efficiency social campaigns are better branded, more involving and more enjoyable than campaigns with a low Social Efficiency score. High Search Efficiency campaigns add being more persuasive to the mix, confirming that successful delivery of a motivating message has sparked active interest in the brand.

While individual results obviously vary by brand the fundamental conclusion must be that if you want people to talk about your brand or search for it you need to make sure your campaign is well branded and engaging, otherwise little is going to happen. But that should not be a surprise should it?