Customer Analytics Summit 2014 Chicago: Day 2 Highlights

Highlights from the presentations by Big Data & Analytics experts from Microsoft, Sears Holdings and Obama for America on day 2 of Customer Analytics Summit 2014.

Customer analytics is becoming critical. Customers are more empowered and connected than ever. And becoming more so. Customers have access to information anywhere, any time – where to shop, what to buy, how much to pay, etc. That makes it increasingly important to predict how customers will behave when interacting with your organization, so you can respond accordingly.

ieanalyticsCustomer Analytics Summit (June 19-20, 2014) was organized by Innovation Enterprise in Chicago. It brought together analytics executives and data scientists working in retail, ecommerce and consumer goods, offering unique insight into the innovations that are revolutionizing their relationship with customers.

We provide here a summary of selected talks along with the key takeaways.

Highlights from Day 1.

Here are highlights from Day 2 (Friday, June 20, 2014):

Jeff HamiltonJeff Hamilton, Head of Consumer Insights, Xpress Browser, Microsoft kicked off the second day of summit delivering a talk on “Emerging Market Insights Combining Big Data & Mobile Market Research”. Browsing on mobile is pleasurable in US due to availability of 4G. However, many markets remain network challenged in delivering a quality internet experience to 4G Smartphone users. Feature phones continue to represent the majority of consumers in emerging markets.

Giving a quick introduction of Xpress Browser, he mentioned that the service uses advanced, cloud-based technology with high speed Internet connections to render and optimize web pages. When user requests a web page using Xpress, the request is routed through a Microsoft Mobile server on its way to the destination website. The Microsoft Mobile server receives the page from the destination website, renders and optimizes it for user’s device, compresses the data and sends the response back to user’s device. This optimization can shrink web pages substantially, and as a result user can experience faster browsing and browse more web pages with your data plan. They also conducted a survey which resulted in these key findings:
  1. Mobile Market Research in not optional when seeking insight into emerging markets.
  2. Emerging market consumers are willing, able and eager to share their needs, wants and desires via mobile.
  3. Unprecedented insight into emerging markets are available now.

Bharat Prasad, Big Data Architect, Sears Holdings talked about “Using Data Mining and Machine Learning in Retail”. He said “Big Data can no longer be defined by the amount of data, but by the type, speed and storage capacity needed to compute and analyze that data.” He stated the problem with large scale data processing is that using traditional computer processing it can be difficult to compute everything, due to storage space, processing time, and cost.

Hadoop brings infinite scalability, extremely large storage capability and fast data processing. Discussing about Big Data Analytics in retail, he mentioned Mahout – an Apache Foundation software project using scalable machine learning algorithms. He briefly discussed three primary algorithms: Clustering, Recommendation Systems, Market Basket Analysis. He illustrated with examples that how these algorithms can be help enhance sales of retail stores. He concluded the talk discussing various layers of Big Data stack. Big Data Stack - Sears
Peter TannerPeter Tanner, Modelling Analyst, Obama for America delivered a talk on “Using Analytics to Help Win the Presidential Election”. He mentioned that the similarities between customer analytics in the Obama Campaign and in the corporate world are quite striking. He mentioned lines from The New York Times dated Oct 13, 2012: “And as the race enters its final month campaign officials increasingly sound like executives from retailers like Target and credit card companies like Capital One, both of which extensively use data to model customers’ habits.” During campaign several large sources of data were joined together. The campaign had 3 main goals for 2012 election: Register, Persuade and Turnout. Statistical models helped during the campaign to achieve these goals. A support model was built to predict who like Obama over Romney and then a turnout model to predict who would actually vote. Turnout Model
He briefly discussed how the team came up with a persuasion model to target undecided likely voters. A lot of fundraising emails were sent. They found that emails subject line “Hey” raised majority of money. He also discussed about Obama Facebook app and polling. Peter concluded the talk predicting the following for 2016 elections:
  • Hilary will run (90% odds)
  • More individually targeted video
  • Email won’t be affective
  • Republicans will catch up in analytics