InfiniteInsight® is not only easier to use and faster at delivering results than traditional data mining tools like SAS and SPSS, InfiniteInsight® also delivers superior models.
Director of Business Intelligence and Billing
Mobilink Boosts Campaign Response Rates Up To 380%
- Pakistan's leading cellular and Blackberry service provider, was facing explosive growth as the market for mobile communications grew 20x from 2004 to 2011.
- Fierce competition among carriers was reducing the average revenue per user (ARPU), increasing customer churn, and eroding profit margins.
- Wanted to use predictive analytics to build customer trust, increase revenue and improve loyalty, but it needed a solution that could make sense of data from nearly 35M subscribers and 200K retailers in 10K cities, towns, and villages.
- Selected InfiniteInsight® for the robustness and agility of its solution, as well as rapid deployment and ease of use.
- Standardized on InfiniteInsight® after examining solutions from other vendors such as SAS and SPSS, because of its excellent predictive capabilities and compatibility with other solutions.
- Applied predictive scores directly in-database to a Teradata environment that is integrated with Informatica and Microstrategy.
- Distilled more than 1,250 variables down to sets of the most relevant and predictive variables for each customer segment and subsegment.
- Deployed new predictive models in less than 1 day.
- Thanks to the in-depth insight provided by InfiniteInsight®, the company revamped its retention strategy and began focusing on preventing churn in addition to predicting it.
- Reduced attrition by identifying customers with a high propensity to churn and taking a “next best activity” approach by offering the right incentive to the right customer at the right time.
- Employed InfiniteInsight® Social for social network analysis to detect both large and small communities of interest in order to implement viral marketing to acquire new customers, segment customers, boost adoption of new products and services, and reduce churn.
- Uptake of customer retention offers has soared eight-fold, from .5% to about 4%, and at a fraction of the cost.
- Quantifiable reduction in churn.
- Determined that 50 percent of potential churners come from 10 percent of subscribers, considerably limiting the total number of customers they need to target in churn campaigns.
- Identify more than 6 million distinct subscriber communities and thousands of influencers or “queen bees”.
- Boosted campaign response rates Up To 380% with social network analysis using InfiniteInsight Social.
- Using InfiniteInsight’s™ Social combined with clustering techniques, Mobilink has been able to discover such insights as:
- 25% of subscribers recharge frequently from their preferred retailer.
- 32% of retailers have a very loyal customer base.
- Achieved “next best activity” strategy – offering the right incentive to the right customer at the right time.
- Better and more targeted promotions and more effective marketing campaigns.
- Consistent engagement of customers and communities in the Mobilink network.
- Increased adoption of value-added services such as SMS, ringtones, music, etc.
- Increased analyst productivity in modeling and preparing analytical data sets.
- Mobilink’s big data analytics approach includes building predictive models from the following:
- 1,250 customer attributes (“columns”)
- 4.3 billion raw monthly call detail records (CDRs) (“rows”)
- 900 million distinct monthly communications (“links”) between 70 million phone numbers (“nodes”)
- 6 million communities of highly connected social groups
- Hundreds of terabytes of data stored in a Teradata data warehouse
- 300 gigabytes of raw data daily (incoming and outgoing CDRs)
- Big data sources including call detail records, product subscriptions, voucher transactions, package conversions, cell site location