ANALYTICS
SECTION ONE
SECTION TWO
 |
Getting Ready for Watson
By Implementing Shopper Analytics
Watson, the IBM computer that handily beat two Jeopardy! champs on
the popular TV game show last year, has a Facebook page. IBM has
moved past such cultural footnotes and is now expanding its Watson
technology. In a recent virtual briefing, two IBM executives explained
where the Watson technology is headed, and how retailers can use
existing insights to “operationalize” customer analytics to create a more
compelling shopping experience.
Cathy Lasser, IBM's vice president and chief technology officer for the distribution sector, explained that IBM is initially targeting other areas than retail, such as healthcare and technical support, to take advantage of Watson’s natural language algorithms. Watson was designed to deal with unstructured data in a question-and-answer format (or answer-and-question in the case of Jeopardy!), so its use in the structured POS data-rich environment of retail and CPG is more limited. But the company is exploring how to enable retail customers with its deep analytics to more easily find what they are looking for, and answer consumer questions, says Craig Silverman, business analytics leader. This could possibly happen by the end of the year.
Silverman cited a global survey of more than 30,000 people in explaining that today’s consumer is smarter, more informed, more technologically enabled, and more demanding. These consumers are defined as instrumented, interconnected and intelligent.
Instrumented Through technology, consumers have instantaneous access to information about retailers, products and other consumers’ experiences.
This “instrumented” consumer is willing to use two or more technologies - such as mobile phones, the internet or kiosks - for researching and purchasing products. Their number rose from 36% in 2010 to 49% this year, with a 92% increase in the use of mobile technologies. The study showed that only 14% of the global population is unwilling to use any technology for shopping.
Interconnected Consumers use multiple technologies to interact with other consumers and with retailers.
While manufacturers and retailers still have influence over consumers’ opinions, this is now a small portion — less than 20% for most categories — of who they trust. Social networks consisting of family and friends, customer reviews and product experts are trusted more as consumers make their purchasing decisions. Most importantly, consumers are in control of where they get this information. Personalized interactions, including the assortment, are in high demand.
Intelligent They have clearly defined expectations of what they want from their retailer now and in the future.
Responding to these consumers involves listening to them, changing the retail organization and improving the shopping process. To do this, retailers need to put analytics in place to listen to the consumer, learn from the content they control, and the factors influencing and motivating them. Companies also need actionable analytics to determine how to change merchandising and marketing to be more responsive. Finally, the retailer needs to empower the consumer to self-interact and chose their preferred channel.
Building on this knowledge, retailers need to operationalize analytics around their customers. This starts with customer insights: Knowing the customers, what they buy, what they are saying in social media, and what they are saying to the retailer. This is the foundation of optimizing and executing decisions across marketing, merchandising and supply chain.
There are three steps to accomplishing this:
- Single View of the Customer: Integrate point-of-sale, loyalty data, e-commerce, social media, and link transactions across channels.
- Customer Segmentation: Develop a deep understanding of customer segment behavior and response to marketing and merchandising activity.
- Analytics Platform: Establish an infrastructure of required hardware and software components for the rapid deployment of analytic solutions.
Modeling the purchase, and the response to retail marketing and merchandising activities, is critical to gaining a better understanding of consumers. This involves moving beyond knowing what they’ve bought and their past purchase patterns, to modeling that behavior and understanding how consumers are responding to what retailers are putting into the marketplace; that is, how the value proposition of the stores is resonating with the consumer, and whether different marketing activities are actually stimulating and driving consumer sales.
After modeling the purchase and response comes the essential step of breaking it down into consumer action clusters, and understanding which part of the customer base is responding to which vehicle or part of the value proposition. This could be assortment, pricing, or something else.
This customer segmentation approach is tailored to each company’s business model, customer data and operational practices. And it yields highly homogeneous, differentiated and actionable customer groups or action clusters.
By analyzing and segmenting the customers, retailers can apply this knowledge to marketing, whether it be media mix, marketing mix, or one-to-one targeted marketing. This can now be done with the knowledge of which customer segments are most desirable to go after, and which approaches will give the highest satisfaction to the customer and the highest return to the retailers.
All of these analytics are about making better decisions. These kinds of analytic tools allow users to run different scenarios, and move investments to affect business and customer groups. As a result of putting the customer at the center, the way retailers operate changes in marketing, merchandising or supply chain allows them to keep up with rapid changes in how consumers shop.
Once that strategy and assessment is done, the next phase is to prioritize and target where the retail organization wants to go, what capabilities it wants to build within marketing and merchandising, and then identify gaps between today’s structure and what is planned for the future. These may be process gaps in the organizational structure, the talent in the organization, or the technology design. Often there will be pieces of all these elements that are likely to be needed for the journey.
The final step is to define a priority list of key initiatives where the highest payout is, and prioritize them, looking at a business case for each, so the retailer will know where it can get the most value, as well as serve its customers better.
Analytics can be very powerful, but it is important to lay out a framework and a plan.
The above article was based on a virtual briefing,“Watson: Today, Tomorrow and the Future of Retail,” conducted by IBM's Retail Industry Practice. For more information, contact Don Gordon at don.gordon@us.ibm.com, or check out the IBM Retail Virtual Briefing Center at ibmwatson.com.
Click on the LinkedIn logo to join the new Shopper Technology Institute Discussion Group
SECTION THREE
