Seattle Times Segments and Scores
March 2006
|
Challenge: Increase subscription and retention rates for daily newspaper
Solution: Use predictive modeling to develop a comprehensive segmentation strategy
Result: The Seattle Times Co. now is able to vary its prospecting and retention messages to different segments
When The Seattle Times Co. (STC) set out to increase the subscriber base for its daily newspaper and get a clear picture of its retention rates, the multimedia publisher knew that good, solid data would be the backbone of such an initiative, so it got busy gathering a robust set of survey-based attitudinal data on both customers and nonsubscribers. But once it had that data, the challenge was finding a way to turn what it knew into what it could use. That’s where predictive modeling came into the picture.
“We wanted to better understand the various acquisition strategies and retention strategies we might undertake. And we needed a way to differentiate among our segments,” recounts STC’s strategic research manager, Janet Farness. “We had developed a segmentation [strategy] apart from our database, based on survey research, and we needed a way to marry our learnings from the survey and then deploy them to the marketing database, so we could use what we’d learned to enhance our results.” With this survey data in one hand and a marketing database full of demographic data in the other, STC reached out to Chicago-based analytics firm Apollo Data Technologies to help guide the correlation process.
As Jeff Kaplan, principal, client services for Apollo, explains, his company took the profiles that STC had built, such as what age range subscribers fall into, newspaper sections they are interested in, their reading habits, etc., and predicted which nonsubscribers fit into those segments.
The resulting segmentation strategy now is being applied to a variety of customer touch points, according to Nadine Selden, a research manager with STC. In October, the publisher dropped a direct mail campaign with versioned messaging for the different segments. Those segments also will be used to track retention rates on the back end, a process Selden expects will take a few months, possibly even a year, to measure fully. However, she can say that up-front response to that campaign is on par with what STC had been hoping it would be. In December, the publisher also initiated a customer service messaging effort based on this segmentation strategy, in which customer service reps are able to deliver specific messages and offers relevant to the segment a caller falls into. “This has been huge for us,” asserts Farness.
Looking forward, STC plans to partner with Apollo once again to undertake a new project: using transactional data to build retention models. “[They’ll] be able to score which customers are likely to churn or cancel,” explains Kaplan. “If you know 45 days in advance that a subscriber is likely to churn, you can call and make them a different offer. It’s always easier to keep a customer than try to get a new one.”
—Tracy A. Gill
Solution: Use predictive modeling to develop a comprehensive segmentation strategy
Result: The Seattle Times Co. now is able to vary its prospecting and retention messages to different segments
When The Seattle Times Co. (STC) set out to increase the subscriber base for its daily newspaper and get a clear picture of its retention rates, the multimedia publisher knew that good, solid data would be the backbone of such an initiative, so it got busy gathering a robust set of survey-based attitudinal data on both customers and nonsubscribers. But once it had that data, the challenge was finding a way to turn what it knew into what it could use. That’s where predictive modeling came into the picture.
“We wanted to better understand the various acquisition strategies and retention strategies we might undertake. And we needed a way to differentiate among our segments,” recounts STC’s strategic research manager, Janet Farness. “We had developed a segmentation [strategy] apart from our database, based on survey research, and we needed a way to marry our learnings from the survey and then deploy them to the marketing database, so we could use what we’d learned to enhance our results.” With this survey data in one hand and a marketing database full of demographic data in the other, STC reached out to Chicago-based analytics firm Apollo Data Technologies to help guide the correlation process.
As Jeff Kaplan, principal, client services for Apollo, explains, his company took the profiles that STC had built, such as what age range subscribers fall into, newspaper sections they are interested in, their reading habits, etc., and predicted which nonsubscribers fit into those segments.
The resulting segmentation strategy now is being applied to a variety of customer touch points, according to Nadine Selden, a research manager with STC. In October, the publisher dropped a direct mail campaign with versioned messaging for the different segments. Those segments also will be used to track retention rates on the back end, a process Selden expects will take a few months, possibly even a year, to measure fully. However, she can say that up-front response to that campaign is on par with what STC had been hoping it would be. In December, the publisher also initiated a customer service messaging effort based on this segmentation strategy, in which customer service reps are able to deliver specific messages and offers relevant to the segment a caller falls into. “This has been huge for us,” asserts Farness.
Looking forward, STC plans to partner with Apollo once again to undertake a new project: using transactional data to build retention models. “[They’ll] be able to score which customers are likely to churn or cancel,” explains Kaplan. “If you know 45 days in advance that a subscriber is likely to churn, you can call and make them a different offer. It’s always easier to keep a customer than try to get a new one.”
—Tracy A. Gill
|
Social Media ROI
Email Marketing that Works (2nd Edition)