Solving Unstructured Versus Structured Data - IBM Pitches Retail Marketers At NRF Show

By
  • Facebook
  • Google Plus
  • Twitter
  • LinkedIn

NRF - The Big ShowA Sunday start to the National Retail Federation's annual "Big Show" brought together the biggest global brands in retail at the Javits Center in New York City.

In a show like this, where the focus is the bottom line and the bottom-of-the-funnel, IBM was again making its pitch to brands and the CMO, in particular, with an IBM-sponsored panel titled "Critical Developments in Retail Marketing: Understanding Consumers, Building Brands."

IBM SVP of Marketing & Communications Jon Iwata, who is also chairman of IBM's strategy team, laid the ground work for the panel with an opening presentation emphasizing themes of "real-time, insights and analytics" in today's marketing.

Overall, today's innovation is not simply about tech according to Iwata, it's about data, and everybody's favorite buzz phrase, "Big Data." Iwata offered up a stat first popularized by Google chairman Eric Schmidt that the equivalent of all the data from 2003 and before is now created every two days. Consequently, another IBM stat: 90% of the world's data was created in the last two years. That's a lotta IBM servers.

So what's the challenge beyond volumes of data? Iwata asked the crowd to consider the three V's:

  • Velocity – There's real-time data flying at us all the time. Marketers must find a way to capture it.
  • Variety – There are two different types of data. At a high-level it's about about structured data (for example, what you might see from an Excel spreadsheet about your company's balance sheet) versus unstructured data types (videos, UCG data, medical images) where there are no hard and fast rules for the attributes of these data types. And not helping things is the fact that, according to Iwata, unstructured data accounts for 80% of the world's data today .
  • Veracity – The accuracy of the data... Verify! Validate! Is it trustworthy? Today's ad verification and emerging audience verification world in display media is a microcosm of this.

So, now we know the challenges, what's a marketer to do with big data. Iwata shared his company's ideas…

  • Segmentation – You gotta cut it up! But, IBM thinks "even today's best segmentation models are blunt instruments." Rather than getting into "spray and pray" mode, it's time to be highly targeted and the marketer's segmentation models need to reflect this.
  • Decisionmaking – Iwata noted its time to move away from using just historical data to ultimately provide insights – "historical" is what IBM says 80% of marketers use today. It's time for real-time and predictive data.
  • Seeing Patterns – And finally, appealing to data-driven digital marketing lovers everywhere, Iwata pointed out the need to look for patterns which provide an opportunity for predictions. Using the example of a large bakery chain which had taken a deep-dive look at its sales data, the bakery found cake was the food of choice on rainy days, grilled sandwiches on hot days, etc. – the bakery could "program" its product line accordingly. Iwata said, "The more data, the smarter the algo becomes." Feed the algo! (my words)

The result of all this segmentation process syncs with today's just-in-time mentality in retail: "Retailers have just what they need, when they need it, personalized and customized per user." To drive the point home, Iwata offered up another quote from Unica's Yuchun Lee, "Good marketing and good retailing is going to feel like a service to the customer." And obviously IBM wants to be at the center of this retailing data storm.

In the panel discussion that followed, real world examples (or "ringers"!) took to the stage as Billy May, VP of ecommerce and Cross Channel Marketing at Abercrombie and Fitch, and Alexis Maybank, Founder and CMO of Gilt Groupe, explained that personalization and segmentation remained the themes of today's successful, digital, data-driven marketer.

Abercrombie & Fitch's May noted that the majority of their customer's "A&F time" is spent online, but they are also a traditional brick and mortar retailer. He emphasized the importance of creating a branding experience that the consumer wants to share. Get that sharing data flowing!

Unlike some traditional retailers who use loyalty cards and incentives, May's focus was on creating lasting relationships with the brand. Why? If a consumer trusts the brand, the more willing they are to give them still more data for targeting purposes – not just name and address, but GPS location, too. Happy customers happily provide data.

Turning from the multi-channel world of Abercrombie & Fitch to Gilt Groupe's online-only presence, Gilt's Alexis Maybank said that they send out 3,000 versions of a daily email to their customers. Testing new marketing initiatives is a constant at Gilt as they track "engagement" (an increasingly nebulous buzz word in marketing), clickthrough rate and all of the related data, which is then poured into the Gilt analytics machine, if you will, looking for a better way to communicate with the consumer.

So what about all that "unstructured data" spewing from Facebook? Can it be captured and made actionable? Iwata claimed it can and that to get the total picture of the consumer, the retailer has no choice – they have to capture and understand what's happening within Facebook's digital walls.

On top of all of this, and echoing something Iwata said at last year's ANA conference, Iwata emphasized that CMOs need to talk and think about the character of their own company in the era of full transparency.

Marketers may get to see more sides to the consumer than ever before with the help of "big data", but it works both ways as the consumer reviews the marketer.

By John Ebbert

  • Facebook
  • Google Plus
  • Twitter
  • LinkedIn

Email This Post Email This Post

By on at

Leave a Reply