Database modeling is the process whereby thousands of variables are mined to identify variables that have high marketing value and are predictive of customer behavior, the behavior that is unique to your customer. Rickard’s data driven modeling solutions may include:
Good Customer Match: Takes your profile to the next level where your customers are matched against a proprietary database to determine which variables are the best predictors for identifying prospects that “look like” your best customers. There are thousands of variables that are factored in. The potential universe of names is scored from the person who is most likely to respond the person who is least likely to respond to your offer.
Regression Models: Takes into consideration all the lists that you are mailing including your house file and other acquisition files. The key difference is that they look at those people that have responded from prior campaigns and those that did not. These files are matched to the Database. A regression model is built to determine which variables are the best predictors for identifying responders. A scoring equation is developed which is used to select names for roll out. A detailed model report, including model variables, gains chart, and ordering recommendations, will be provided upon completion of the build.
Optimization Models: The response rates for direct mail vary by offer but, at the end of the day, the majority of mail goes unread. The postal costs are high so optimization models help determine the best prospects by scoring the net mailed file and suppressing the bottom 10-15% who are least likely to respond. By applying this custom model to rank the consumers in your campaign for expected response and profitability. You select the consumers to be suppressed from your campaign ― across sources ― and save mailing costs for non-responders or poor payers helping you to mail more efficiently
Re-Activation Models: Most clients keep mailing their lapsed/former/expired clients over and over again. The Reactivation models allow us to figure out the best vertical and tag your former customers with current transaction information. If we know they are buying then they are more likely to respond than someone who is not currently transacting in any way.
Audience Profiling
You think you know your customers but do you really? We often find out something new about who you think your customer is. Profiling and Modeling goes beyond the static demographic and gets into what they like to do, what they spend their money on, how they like to be contacted and so many other variables.
Case Study: Large brand in the mobile phone sector was spending their budgets targeting the young, male, hip, 22-28 year old technophiles who needed the latest and greatest. It turned out after profiling their best customers that their target audience was much older and actually less technically savvy than anticipated. This new revelation prompted a change in messaging and an increased traffic to the individual carriers.
Lead Scoring
Rickard can use modeling, analytics and third party data to score your sales or CRM database for your sales management team. The end result is qualified and prioritized leads based on modeled firmographics, behaviors, transactional history, propensities and more. Qualified and prioritized leads means a happier salesforce, more efficient campaigns, higher conversion rates and a stronger ROI.
