"Where is our conversation going? What are the little indicators that say 'it's not going in the right direction?' What are the little indicators that tell us early there are opportunities to steer it where we'd like it to go?"

In 1996, Don Peppers and Martha Rogers published The One-to-One Future. The book posited that by accumulating data, it would be possible to build relationships that could give customers a fully tailored experience. That hasn't happened. The behavior of shopping has changed drastically. The majority of customer interactions moved from face-to-face to digital, requiring businesses to look for other ways to determine who the customer is and what the customer's needs are. The technology has not advanced enough to do this. The change is continuing, and the technology isn't keeping up. As an example, it's not enough to just send a message to someone walking by your store reminding them to come in. The GPS monitoring needs to see whether they're coming or going, and other aggregated data needs to look at their shopping habits and present them with personally-tailored offers. Another example: Instead of a car dealership just telling customers when their car next needs service based on the odometer, the dealership will be able to predict, based on travel patterns, when the car will need service, where a convenient place to drop the car off is, when a good time to drop the car off is – create a provisional appointment which can be confirmed or declined nearly automatically- , and will be able to provide a replacement car for the time of the service, to create a nearly seamless experience for the customer, which gives him incentive to use the service regularly and on time. Technology has grown in incredible ways, meaning that companies will soon be able to predict a customer's needs.

Two fundamental problems with what we have today:

1. Our view of the customer: In 1996, most interaction was face-to-face, by mail, and in call centers. Ten years later, nearly half of the total marketing influence is online. The two systems (old and new) often do not exchange information, giving an incomplete view of the customer. In the past 5 years, that view has fragmented even further, due to mobile and different communications platforms. Connecting the different channels is crucial to understanding and predicting customer behavior.
2. There is too much information: There needs to be a mechanism which distills actionable items from the mass of data. The goal needs to be to aggregate all of the data, all of the events, so that we can predict and affect future actions. The data structure must preserve all of the events, in time order, to be able to understand the connected journeys of customers.

The Causata system is called Digital Customer Intelligence and it aggregates all the data and helps distill the data down to actionable items.

Paul Phillips - COO , Causata

Paul is the CEO and Founder of Causata, a software company that deploys a new type of customer intelligence software for Fortune 1000 companies. Paul is a pioneer in helping companies to win by becoming more data driven. He's also an experienced entrepreneur with a track record for building world-class companies. In 2000 he founded Touch Clarity, a software company that developed the online content targeting platform that was to become the platform of choice for the largest consumer Financial Service companies, including American Express, Wells Fargo, and Bank of America. Following the acquisition of Touch Clarity, Paul worked as VP Test & Target at Omniture Inc. until 2008. Prior to 2000 he was a Managing Director for Urban Science, a retail network planning and database marketing company with an outstanding data mining team that won the KDD Cup twice. Paul is a frequent speaker at conferences about marketing, customer experience and analytics. He also maintains a close relationship with the academic community - most recently as part of the Strategic Advisory Council for PASCAL, an initiative for advancing the sharing of knowledge relating to pattern analysis and statistical computational learning. Paul holds a BSc from Bristol University.