Gaining sales is good, but finding profitable customers who stay committed is even more important in rapidly changing fields such as satellite TV, where individuals can be quick to hop from one service to another.
With this in mind, Southern Pines, N.C.-based Trident Marketing's technology leaders turned to an advanced analytics engine that provided quick and useful measures of consumers' likely staying power. The goal was to effectively reduce the risk of what is called "churn," or customer turnover.
"A natural problem we had to solve was to keep the customers [that our clients already] had," said Brandon Brown, who is Trident's chief information officer (CIO). "We wanted to be able to leverage analytics to help lower our churn numbers. In any business, you are going to want to do that."
For this effort, Trident implemented an IBM Netezza
Using in-database analytics to target customer churn
While many benefits were gained, the first objective was to lower churn rates for Trident's customers. Trident is a direct response marketing and sales firm that operates call centers, promoting brands like ADT, Travel Resorts of America and DirectTV. To better sell products for companies like these, Trident leaders felt they needed to improve chances that customers would stay with their offerings long after the sale.
If you are not investigating the potential uses of analytics in your company right now, then I hope you are writing your resume.
Brandon Brown, CIO, Trident Marketing
That required changes in its approach to data management, according to Brown. It meant moving analytics closer to where the data was and making results available to call center operators in real time. With in-database analytics, Trident telemarketers can now quickly pinpoint which geographic markets to target, as well as who and when they should be called.
A key benefit of the Netezza database appliance, said Brown, is that it "can download an analytic model onto the Netezza box and have it run close to the data." That provides better performance and easier implementation than approaches requiring one to move data around a lot. This removes what Brown calls "a whole layer of extract, transform and load (ETL)" work.
"We wanted customers that would not churn and that would keep their equipment for the long term. We found the game changer in all this was our ability to process all our analytical data in real time. That never happened before," said Brown. To build the analytical model and libraries that ran directly on the Netezza box, Trident turned to Fuzzy Logix LLC, a Charlotte, N.C.-based provider of analytics software and professional services.
The analytic models can mix data from customer relationship management systems and order entry systems with external data from credit management services and search engines. Running the models on the Netezza appliance enables Trident to better predict people's inclination to quickly cancel services. Similar analytics allow Trident marketing department staffs to more efficiently select appropriate keyword phrases for sponsored search engine links, cutting costs for advertising.
Analyze this -- or write your resume
Brown cautions that IT leaders considering similar efforts should work closely with other departments to set project objectives.
The analytics work can display its worth in a number of ways, he emphasized, but it entails significant expenditures for equipment and services to get up and running. Therefore, he said, it's important to keep company executives and project stakeholders informed about the initiative’s forecasted return on investment. "You better show that within X number of months you can return profitability on that investment," he said.
More on analytics
Get a ringside seat for data warehouse appliance slugfest
Learn how analytical capabilities set innovators apart
Check out Forrester's view on predictive analytics
According to Brown, the new system at Trident helped his IT department drive a better relationship with the chief marketing officer and the chief financial officer. "When we did analysis, we thought it would be 18 to 24 months before we could see a return on the investment made," he said.
Within eight months, better estimation of promotional campaign costs freed up capital reserves, putting more money back into operations' cash flow, he said.
"I would never have promised that, but that's what happened," Brown said. The ability to push better information into the sales process via in-memory database analytics has also improved churn rates and helped return gross profit to the bottom line.
Brown counts analytics itself as a game changer. He has been in "the IT business" for 25 years and sees today's data analytics software tools as vital. Fellow CIOs need to keep up, or face consequences.
"I would tell any group of CIOs, 'if you are not investigating the potential uses of analytics in your company right now, then I hope you are writing your resume,'" he said.