Current Issue
This Month's Issue

Follow Fast Company

We’ll come to you.

5 minute read

Analytics And Info Overload: Insights From eBay, Adobe, And Oracle

Are analytics a prescription for information overload? Three top Silicon Valley tech titans share views on what's new in e-commerce and retailing.

Last week I moderated a panel for FountainBlue which explored the latest trends and opportunities in e-Commerce and Retail Analytics. This panel included three distinguished executive panelists from Oracle, eBay and Adobe, namely:

Tobin Gilman, VP of BI & Performance Management Product Marketing, Oracle

Darren Bruntz, Senior Director, Analytics Platform & Delivery, eBay

Raj Sen, Group Manager, Multi-Channel Analytics, Adobe

It was a rich conversation so I thought I'd share a few key insights.

Analytics Now in the Hands of Many: Once the domain of highly trained corporate retail analysts, access to data and tools are being given to store managers to operate their store location better. Tobin Gilman of Oracle reports, "Store managers have not been traditionally trained to be analysts." Retailers need to make analytics easier to understand and less likely to be misinterpreted. An interesting application of Business Intelligence (BI) in retail is using analytics to measure training effectiveness: According to Tobin:

"There is high employee turnover in retail. Many workers are temporary or seasonal. Training them on job skills and the products they sell is a big investment retailers make. Employee training makes a big difference in retail sales performance and customer satisfaction. Analytics can help them measure the effectiveness of their training programs, allowing them to better target training initiatives."

Panelists also noted that consumers are demanding easy access to peer reviews and rating metrics in addition to readily available product information. Retailers and e-tailers must serve up and enable access to this information in a digestible way without overwhelming the consumer.

Data-Driven e-Commerce: According to Darren Bruntz, "Almost everything is analytically derived, either directly or indirectly, on the eBay Web page today." This includes placement, layout and the personalized user experience for the 5 to 6 million visitors to the site per day. This is a vast difference from a decade ago when there was much less analytic rigor and available behavioral data.

Adobe and Oracle provide advanced tools for retailers and e-tailers to analyze behavioral data and provide a more personalized experience. The panelists stated that today's tools are self-learning and make the experience much more personalized for each user. In other words, rather than static data mining, the programs respond to data gathered from user clicks and behavior serving up results based on an ever-expanding base of knowledge about the customer.

The Multichannel Experience is Increasing Demand for Analytics: As retailers and consumers utilize multiple channels (e.g. mobile, Internet, retail store), the available data to analyze is exploding. This puts pressure on retailers and e-tailers to scale IT and tools quickly to accommodate the data. Raj Sen described how many of Adobe's customers have:

"Terabytes of data from multiple channels and data sources to manage and analyze. Utilizing analytics solutions that balance speed, scale and flexibility to empower you to be able to make quick customer-focused business decisions, is imperative."

The group discussed how cloud computing will enable companies to add capacity when needed to support this explosive growth. The abundance of data also creates greater risk of developing "that's interesting to know" reports instead of "actionable analytics." Providing the right analytics and tools is an escalating challenge.

More Creative Metrics and Visualization Needed: Retailers need to keep up with managing their merchandise online and offline. Along with the increasing volumes of data, representing the data in ways that are digestible using techniques such as waterfalls, infographics, and other advanced data representation are important. Such solutions must address the intersection of data on behavior, multiple channels, customer demographics, transactions and products and translate these factors into actionable correlations.

Reducing Transaction Friction: The need to reduce transaction steps for the user is imperative. Customers have limited time and patience with less than 3 % of waking hours devoted to shopping and only 6 minutes a day on average spent on e-commerce sites.

As an example of how technology can help reduce friction, eBay envisions enabling sellers to take a photo of an item from their mobile phone. The system will have intelligence to identify the item and provide a suggested price based on similar items, and then automatically upload and merchandise the product. Other methods such as mobile wallets were also seen by panelists as ways to make transactions faster and simpler for the customer.

Need for Real-Time Analytics: The majority of event attendees noted that they had use mobile price comparison apps such as RedLaser while shopping. These new behaviors in the store require more real-time responsiveness and promotions. In addition, social shopping and social CRM provide additional dimensions to the experience responding to customer questions and complaints in a timely manner is expected.

Tension Between Personalization and Privacy: All agreed that personalization is central to a more relevant and optimal user experience. In some cases, technology doesn't personalize enough. For example a Netflix account shared by a family may be biased toward teen action movie recommendations because the teen uses the account more heavily; Netflix cannot discern between different family member preferences.

However the panel also recognized that not all users will have a personalized experience, "We will never get to 100 % of users because of privacy concerns" It was noted that some users purposely don't want their information to be known and will "fake" the system to mask their identity. This presents a "chicken and egg" syndrome making it difficult to provide a relevant experience if the preferences of the user are unknown. To remedy this issue, retailers must develop a bridge of trust with the customer about data protection and security.

Several people in the audience pointed out that they were not too concerned about sharing their information with a particular e-tailer to get a more relevant experience, but that there was concern over where else their information and profile would be shared. This knowledge should provide opportunities for retailers and e-tailers who create a strong policy of trust.

Actionable Analytics

No doubt the role of analytics is changing dramatically with multichannel methods, advanced data representation techniques, and social features. A key challenge in the future will be designing relevant filters, measures and reports that keep pace with the explosion of data and differentiate between actionable and "nice to know" information both for experienced analysts as well as the lay person. I look forward to witnessing how this exciting field evolves.

Adrian Ott is the award-winning author of The 24-Hour Customer: New Rules for Winning in a Time-Starved, Always-Connected Economy which was named a Best Business Book 2010 by Library Journal and by Small Business Trends. She is also CEO of Exponential Edge Inc. and a frequent keynote speaker. Follow Adrian on Twitter at @ExponentialEdge.

©2011 Exponential Edge Inc., All Rights Reserved