Leaning forward in data conversations

Columbus general manager, Graham Wilkinson
By Columbus general manager, Graham Wilkinson | 20 January 2017
Graham Wilkinson

Graham Wilkinson, general manager (VIC & WA) for Columbus, looks at the divisive topic of data – and how to make it more useful.

Lately I’ve been thinking about the rise of big data, and the many ways it has both shaped and fragmented our industry. Programmatic has given marketers the license to talk about data incessantly, regardless of the fact that a majority of us don’t have formal training or qualifications in this space.

For the last 3 or 4 years, data has been a widely discussed and often divisive topic around the world, in part thanks to Edward Snowden and Julian Assange. During this time I’ve been involved in, as I’m sure many of you have, countless meetings discussing data washing, buying, stitching, binding, matching, modelling – the list goes on. To add to the complexity, the source of this data can be equally baffling – 1st party, 3rd party, site-centric, ad-centric, etc.

This fundamental evolution of marketing has spawned a fragmented industry – one full of businesses that claim to make data more accessible, but very rarely do they make it more useful.

The Untapped Data Goldmine

To circle back to the beginning, it dawned on me that although search marketers have been privy to many of these data-centric conversations, we have been guilty of not really getting involved or adding true value. It’s the nature of any search marketer to optimise and create efficiency, which requires forensic investigation at a granular level. This is great when your job is to drive down a brand’s cost-per-acquisition, or increase return-on-investment, but it can often be a debilitating trait to have when looking for macro trends.

The reality is that search and performance marketers are sitting on a data goldmine that is not being tapped into from a wider behavioural, media and marketing sense. We need to sit forward in conversations about data and start to add real value. One simple way we can do this is by using the free Google Trends tool to gain behavioural insight into major events. Primarily, this tool is used to map overall search volume at a brand or category level to justify increased investment into search. In actuality, Google Trends can do a whole lot more than that.

The Proof Point

The team at Columbus recently conducted some analysis into the impact that the Olympics has on search behaviour during the event. We wanted to test the hypothesis that consumers only have finite levels of interest; that when they are distracted by a major global event like the Olympics, it acts to monopolise their attention and take them away from their everyday tasks.

We looked at search volumes for 50 of Australia’s top brands across a breadth of verticals, including automotive, finance, FMCG and travel. We analysed three weeks of search data in total, spanning before, during and post the Olympic period. We then compared the data for the same period, across a number of non-Olympic years.

Attention is Finite

During a three week Olympics, search volume for these brands decreased on average by 2.6% compared to the same time period in non-Olympic years. Additionally, the three weeks following the Olympics showed an increased average search volume of 5.4% when compared to the same time period in a non-Olympic year.

Although this study is by no means conclusive, it gives some valuable insight into behaviour during a major event like the Olympics. Contrary to what Google may tell us, this study indicates that there is only a finite amount of search volume available. More broadly, there is only so much attention we can look to capture at any given time.

As human beings, we can’t feasibly be searching for the latest result in the 100m Olympic sprint and at the same time be looking to renew our health insurance. This phenomenon is not exclusive to the Olympics - we see similar trends when a new iPhone is launched or more recently during the American elections. These ‘black holes’ seem to appear, suck a huge amount of attention volume inside, and then vanish as quickly as they appeared.

Develop Your Own Hypotheses

As an industry, it is important to realise that our futures lie well beyond the bounds of search and performance and much more in the realms of predictive analytics, behavioural analysis and data driven marketing.

My challenge to anyone reading this is to put Google Trends or data in general, to the test.
Develop your own hypotheses, test them and start to lean forward in broader data conversations.

By Columbus general manager, Graham Wilkinson

comments powered by Disqus