If a picture is worth a thousand words, imagine how much you can learn about your customers, and how you can keep them happy.

By Naveen Sharma, Xerox Research Center Webster, New York

Naveen Sharma

Naveen Sharma, chief innovation officer for Retail at Xerox.

Xerox scientists are showing up in unexpected places. For example, imaging and data scientists are developing applications that can help retail stores use the data they already have.

Consumer and retail analytics is a growing segment of Xerox’s big data research and it’s not just the pure data they are after, but the hidden value of the data. Department stores, fast food restaurants and drug store chains want to predict consumer behaviors — for example:

  • Where shoppers tend to linger in the store.
  • How customers interact with merchandise or sales associates.
  • Is it easy for customers to find what they’re looking for?
  • Are your customers happy?

Customer analytics unlocks insights that, when applied, can win new consumers, gain consumer loyalty and help retailers grow revenue. Many in the industry say that without applying this type of data analytics, the information is meaningless. Here’s a look at how imaging and data scientists at Xerox can answer these questions:

The Other Reason for Video Surveillance: Most stores have video surveillance cameras mainly for anti-theft purposes. However, that same video can contain valuable information about consumer habits. Where do customers linger most? How are they interacting with merchandise or staff? Are they happy or not? What can the video tell us about emerging trends? Are they looking for something and not finding it easily?

Reviewing traffic flow patterns could result in re-organizing displays to attract more attention. The result could be more purchases, or new processes that make drive-thru lanes more efficient.

Video also offers up cues to a shopper’s emotional state: Xerox researchers have developed algorithms that can detect and determine a customer’s emotional state. Computer vision software and techniques enable machines to “read” a person’s facial expressions and cues, and determine, for instance, is he happy, frustrated, or angry. It can also track what attracts the most attention. Retailers can use this information to rearrange displays or customer services processes accordingly.

Data fusion uses multiple sources: Social media is a great source of data and information that companies may use to improve customer service. However, many social media listening tools rely on keywords to determine sentiment, which isn’t always accurate.

Our researchers use “data fusion” – the ability to combine data coming from a variety of sources and teach computers to determine the sentiment of a comment accurately. The tool analyzes what is said in the context of the conversation, which means it can detect sarcasm, and understand abbreviated words in a tweet or Facebook post. Our clients can use this tool to get at the data quickly and efficiently, so they can be proactive on what could become negative customer service issues – in some cases reaching out directly to that person before they file a complaint. Large customers with call centers, can use this data to prepare call center employees better.

Naveen Sharma is chief innovation officer for Retail at Xerox. He discusses big data and how companies bring innovative products and services to market: Read this five-part interview on the One Million by One Million Blog.