Portrait Software announced that uplift modeling, the basis for its Uplift Optimizer solution, was the driving force behind the award winning submission to MineThatData ’s E-Mail Analytics and Data Mining Challenge. In the winning submission, Nicholas Radcliffe, mathematician and founder of Stochastic Solutions Limited, applied uplift modeling to a dataset of 64,000 customers to determine which email campaign – a men’s mailing or a women’s mailing – had the most impact. Portrait’s uplift modeling solution went beyond responding to the initial challenge and was able to call out those in the targeted group that were best left alone so as not to create a negative effect from the marketing and potentially make the customer less responsive.
“Being able to eliminate negative effects from marketing campaigns is critical. E-mail marketing is often considered to be effectively free, but only because no one currently counts the cost of such negative effects,” said Mark Smith, executive vice president at Portrait Software. “Marketers often do not realize that their marketing efforts can sometimes drive people away or prompt people who would have purchased to look for competitive offerings. Uplift modeling is able to separate these individuals so that marketers can focus on only those who will be positively impacted by the marketing campaign. This translates to millions of saved marketing dollars as well as increased return on marketing investment. Clearly, uplift modeling is a powerful approach and as a recent report by Gartner indicates, it should be reviewed by all marketers looking to build a competitive edge.”
The Challenge, the results
Kevin Hillstrom, through his MineThatData blog, made available a dataset of 64,000 records, each describing a customer. One third of these customers were randomly chosen to receive an e-mail (referred to as the “Men’s e-mail”); a second random third received a different e-mail (the “Women’s email”); and the remaining customers served as a control group, receiving neither e-mail. The challenge was to analyze the dataset and determine which group was most impacted by the direct marketing campaign or, whose behavior changed positively as a result of the marketing outreach.
Uplift modeling enables highly targeted campaigns
Unlike other more traditional response models, during the course of the challenge Uplift modeling identified customer segments for which the Women’s e-mail reduced rather than increased spending. It was also observed that while average spend among purchasers increased significantly with the Women’s campaign, it declined slightly for recipients of the Men’s campaign.
Uplift modeling – maximizing incremental lift
Uplift modeling predicts the incremental impact – or uplift – a marketing campaign has on a customer and differentiates “The Persuadables” – those that buy but would not have done so had they not been recipients of a marketing campaign - from “The Do-Not-Disturbs” – the group of people for whom the campaign triggers a negative response. Uplift modeling measures the variation in the difference between a treated group and a control group.
Sometimes discussed using different names – differential response analysis, incremental impact modeling, net modeling, incremental response modeling and "true" response modeling – uplift modeling separates "true" responses to campaigns from purchases that would have likely to have happen anyway.
Uplift Optimizer, Portrait Software’s uplift modeling solution, is currently in use by marketing organizations across the telecommunications, retail and financial services industries. It combines a powerful algorithm and an automated analytic engine into a complete solution ready to deploy as an integrated extension of existing analytics environments or as a standalone capability. Uplift Optimizer has successfully demonstrated its capabilities ensuring optimal customer targeting for all forms of customer campaigns including customer retention, up-sell and cross-sell offers. Organizations using Uplift Optimizer include Fidelity Investments, Nationwide Insurance, Lloyds TSB, Telenor and Fingerhut.
About Portrait Software
Portrait Software (portraitsoftware.com) enables B2C companies to drive sales, reduce customer attrition and optimize marketing spend with a next-generation marketing suite. Our integrated solutions empower marketers to automate engagement with targeted individuals and, strategically map delivery of the right message, at the right time, across all marketing and service channels. The Portrait suite seamlessly integrates leading-edge yet easy-to-use analytics enabling powerful, event-driven- and highly personalized customer campaigns. Customer-focused organizations benefit from being able to conduct two-way, intelligent dialog with individual customers, and to build stronger and more profitable customer relationships while maximizing ROI from customer-focused programs.
Our 300 + customers include industry-leading organizations in customer-intensive sectors. They include Merrill Lynch, US Bank, Lloyds TSB, Nationwide, T-Mobile, Telenor, Ziff Davis, Fingerhut, Bank of Ireland, Bank of Tokyo and Fiserv CBS Worldwide.