Apollo Data Technologies plans to launch a predictive analytics solution this week that puts complex statistical tools in the hands of marketers.
“Marketers don’t have time to do anything with all the data they’re collecting, but they don’t want to throw it away either,” said Jeff Kaplan, principal, client services, at Apollo Data Technologies. “Really, 90 percent of data is just noise — it’s the last 10 percent that’s worth examining.”
Predictive analytics mines transactional, behavioral, call center, and geo-demographic data to identify consumer patterns, segments, and behavior. It is, for example, part of behavioral targeting solutions. In marketing, the technology can apply various algorithms to data to understand consumer buying behavior. It’s used for sales and inventory forecasting, retention modeling, and ad targeting.
Apollo has used predictive analytics for such data mining and data warehousing tasks as enabling The Seattle Times to target marketing efforts to various customer segments and helping Barnes & Noble maintain retail store inventory levels. It also applies the technology to a text mining program.
Classmates.com used Apollo’s technology to aid in customer communications. The company once manually tagged all incoming email. This led to inconsistent tags due to differing category interpretations, as well as simple mistakes. The lack of an automated process prevented the company from prioritizing its responses to potential paying customers.
“When people used the pull-down menu with contact options, everybody would choose ‘other.’ There was no queue process, so people that were trying to become paying members were put in the same category as those that forgot the password of their free account,” Kaplan said.
Customers no longer choose a category. A single entry form is analyzed by the text mining engine, which determines the queue a message is placed in and gives it a priority.
“Predictive analytics can mean different things to different people. It can be applied to different pain points and be effective — from customer segmentation to supply chain optimization,” said Martin Schneider, analyst at The 451 Group. “Apollo has taken the approach of creating simple dashboards to make these tools available to non-technical business users and put predictive analytics in an actionable framework.”
Apollo sells its software as a service, with annual subscription pricing. This way, the company makes sure the customer can use what Apollo is trying to sell them, Kaplan said. Before starting the company, he was a product marketer and cofounder Paul Bradley was a developer at digiMine, a data mining company that changed its focus and is now known as Revenue Science.