Big data, budgets and building predictive analytics

In the big data era, the brands that succeed will be those that know what a customer needs before he or she even starts looking for it. It is the possibility offered by the massive amounts of information customers reveal about their needs and desires; this, combined with analytics tools, will allow companies to reach customers before customers even begin looking for a new product or service.

“I think that search is going to be with us for an eternity but is going to be less of a driver in the future,” former Commonwealth Bank CMO Andy Lark, now the CEO of Group Lark, told CeBIT’s Datacom big data conference today.

“The great brands will get to you before you know or need or start searching for a product. We are going to use those enormous data sets to reach you well in advance of the search starting.”

Lark cited one of the more popular examples of this in action: US retail chain Target’s use of data mining to predict whether a customer is pregnant, and bombarding them with relevant offers.

Big data should not be viewed as a “science experiment”, Lark said; it should been seen as a way for companies to “tune and adapt” to their customers in real time. “Your ability for instance to capture the data stream from a domain.com or realestate.com, marry that cookie data with your own big data set,” so when a customer hits a bank’s website on a Monday looking for a home loan, the site they visit will reflect their needs without them need to search for relevant information.

But big data isn’t just shaking up businesses’ relationships with their customers. Because it represents such a tectonic shift, big data will affect how marketers go about carrying out their function within the business.

“Generally speaking, we’ve been able to shoe horn the technology from prior eras into our budgets,” Lark said. When search was young, marketers could dedicate a small percentage of their budget to it, and then increase it over successive years.

“You just sort of squeeze it in, and squeeze stuff out,” Lark said. But in the new world, this “simply doesn’t work”.

“The rate of change is so exponential that the technologies we’re having to implement are just exploding beyond the budget curve. So when we talk about big data, what we’re really having to deal with is not just big data - we’re having to deal with mobile, we’re having to deal with social, we’re having having to deal with the emergence of sensors. And we’re actually having to deal with a significant structural rebalancing of the PNL.”

“I think most business today are going to be very very challenged by this because they’re going to look at their PNLs and say ‘I can’t just shoehorn this big data thing in’, right?” Lark said. “‘I’m actually going to have to look at the full budget stake I’ve got and actually challenge the way PNL operates.’”

“The reason for that is... this can’t be viewed through the lens of ROI. If you try to view the current initiatives ... through the ROI lens it’s going to take you 20 years to get there,” Lark said.