Richard Snow's Analyst Perspectives

PBBI Provides Portrait of Customers through Predictive Analytics

Posted by Richard Snow on Apr 13, 2011 9:25:21 PM

Since last summer, Portrait Software has been part of the Pitney Bowes Business Insight (PBBI) subsidiary. Since the acquisition the combined teams have been putting together a comprehensive set of products to support data quality and customer interaction management. The suite includes the Portrait Self-Service Analytics, Miner, Uplift, Uplift Optimizer, Dialogue, Interaction Optimizer and Foundation modules. The first four provide insights to understand customer interactions and the other three support acting on those insights. Recently PBBI announced updates to both Miner version 6.0 and Uplift Optimizer.

Portrait Miner evolved from Portrait’s original customer analytics product. It has added capabilities to include more data sources, provide 3-D visualization of the outputs and simplify creating models, including predictive models that help users better understand customer behaviors and propensities. Miner’s new user interface guides conversion of complex data into visual information, including 3-D views of customer segmentation, transactional data, geographic analysis and website modeling. It provides several preloaded modeling techniques including decision trees, regressions, additive scorecards and clustering. Automation of models enables companies to address important business issues such as customer acquisition, churn, campaign response, cross- and up-sales opportunities, risk and customer lifetime value and profitability. Using some of PBBI’s other capabilities, users also can incorporate location information.

Most data mining or predictive analytics tools incorporate the concept of lift, in which models predict the responses of segments in a population. Portrait Uplift takes lift one step further. It places all customers into one of four categories – persuadables, lost causes, sure things and sleeping dogs. It uses complex modeling techniques to identify the likely reactions of customers to different marketing actions – for example, what might make a persuadable actually buy something. It also identifies actions likely to have negative impacts – for example, one that might “wake up a sleeping dog” who as a result stops being a customer. This feature can help companies target their marketing actions more sharply, improve returns on their investments and minimize negative outcomes. The user interface guides users in creating models and suggests steps to take, which is useful for the many people who don’t fully understand the complexities of uplift modeling. It is integrated with SAS analytics software so SAS users can control the whole process. It is also integrated with Microsoft Office, which means users can work in a familiar environment and perhaps speed up analysis and decision-making.

This last point is important because our latest research into customer analytics shows that the majority of companies still use spreadsheet software such as Microsoft Excel to produce their customer analyses; for products such as those offered by PBBI to be more widely adopted they will have to be easy to use while still supporting advanced capabilities. In fact, the results of this research could be disheartening to these vendors. They show that the underlying reason spreadsheets are still so widely used is that users are content to maintain the status quo and settle for basic customer-related metrics such as sales values, customer satisfaction scores and call volumes. We at Ventana Research believe that this has to change. Business is now less about acquiring new customers and more a case of retaining them and selling them more products and services, and this requires making it easier for them to interact with a company and delivering better experiences to them – all while maintaining affordability. This effort cannot succeed without complex business-related metrics, and companies need to generate them rapidly so the customer doesn’t leave before the correct action is identified and executed. Suites of products such as PBBI’s can help with these tasks. We recommend that companies review their portfolios of customer-related metrics and consider how products such as Portrait Customer Interaction Suite can help them improve business performance.


Richard Snow – VP & Research Director

Topics: Predictive Analytics, Social Media, Customer Analytics, Customer Experience, Operational Performance Management (OPM), Business Analytics, Business Collaboration, Cloud Computing, Call Center, Contact Center, CRM, Customer Performance Management (CPM), Data Mining, Sales Performance Management (SPM)

Richard Snow

Written by Richard Snow

Richard leads Ventana Research’s Customer and Contact Center Performance Management research practice, which is dedicated to helping organizations improve the efficiency and effectiveness of managing their customers, throughout their lifetime and across all touch points, including the contact center. He conducts research exploring the people, process, information and technology issues behind customer operations management, contact center management, and customer experience management. He also works with senior business operations and IT managers to ensure that companies get the best performance from today’s highly complex application products. Richard has worked in management and consulting leadership positions in the technology industry including with Price Waterhouse, Sema Group and Valors. In his work, he has been involved with all aspects of delivering highly complex IT solutions to a variety of clients in the telecommunications, financial services and public sectors. Richard has specialized in delivering customer care and billing solutions for telecommunications operators, and several multi-channel contact centers for organizations in both the public and private sectors.