Instead, big data analytics today work narrowly from the ground up, addressing issues in the digital-only world, typically around maximizing efficiency of digital campaigns -- for example, how to deliver customized ads up to specific, narrowly targeted, digital-only groups of customers, based on their previous digital behavior.
We need to develop a top-down approach, built to aid marketers with their No. 1 priority, which is to focus on how consumers relate to brands, and how we manage that relationship on an ongoing basis.
If we look at the two big buckets of big-data usage, each is excellent in focusing on a narrow scope, but neither scales up to a holistic marketing solution.
Live data dashboard reporting. The basic level of big-data analysis, live consumer-response data (sales, calls, traffic, website metrics, social data, keywords, etc.) is streamed together in one flexible report. This is a critical function, but not necessarily insightful or predictive from an overall marketing perspective.
Cookie matching. Customers' digital data, from website usage to everything else they do on the web, is collected by matching cookies or tags placed on consumers' devices from various online destinations. A powerful approach to tracking consumers' online path to purchase, this data lacks scale (it covers only online consumers, excluding, for example, those who opt out of online tracking and those who purchase other ways) and accuracy (when people log onto the same internet destinations from multiple devices, it is hard to figure out who is a unique visitor).
How can we make big data address the needs that our clients have every day, to influence overall consumer behavior to create more profitable brand sales?
First, we need to focus on marketing and consumer goals, rather than data-collection efficiency. The key questions we must ask ourselves in launching a big-data analysis should concern which consumer behaviors we are trying to affect. The marriage of a strategic view of the advertising task to the big-data analysis requires a new methodology -- finding ways to blend established consumer-insight techniques with big-data analytics.
Our firm's focus is to deliver new insights and answers to the questions our clients have rather than change the questions to fit the data available. For example, one of our key marketing tasks is to map out the impact for individual brands of each stage of the consumer-decision journey, from awareness to intent to advocacy. By combining the wealth of data we get from consumers on the web, in-store and from responses to media, big-data analytics has helped us understand changes to consumers' brand attitudes in real time and respond immediately.
Take consumer awareness -- it plays a critical role in nearly every campaign, but we know how it really shifted only weeks after the campaign ends. Big -data modeling techniques allow us to define awareness shifts minute by minute, with the multitude of consumer-action data feeds we receive.
By its nature, big data is a live, ever-changing beast. There is no single answer that emerges from the analytics, but rather a flow of consumer-behavior information. The output systems we build from big data should not be a snapshot of a target or a segment, but a guide for engaging customers and prospects in a dynamic, live fashion, with our messages and delivery constantly evolving in response to real-time consumer behavior.
There's no doubt that the ability to interpret and act on big-data analytics represents the future of media planning and buying. But we can't forget the task set by our clients -- to focus on consumers overall and not just their digital selves.