Grameenphone reduces churn to grow revenue in saturated market
With nearly 40 million mobile phone subscribers that account for 42.7 percent of the SIM card market, Grameenphone is the leading provider in Bangladesh. But the company was faced with tough market conditions in 2012.
More than 95 percent of subscribers are from the prepaid market. Average revenue per user (ARPU) is low at less than US$4, and there is a moderately high rate of churn. Business subscriptions account for less than 3 percent of the total user base.
Asif M.K. Bashar Khan, Deputy General Manager of BI Planning and Systems at Grameenphone, says, "In Bangladesh, mobile penetration in the more affluent, urban areas has almost reached saturation. Any new acquisitions are more likely to come from the rural areas where household income is comparatively low."
Grameenphone's big advantage over the competition has been network coverage and quality, "but that advantage is slowly eroding, especially in urban areas, as the competition is catching up," Khan says.
As new suppliers have entered the market, subscribers have learned to shop around for the best deals.
Identifying new revenue streams
Gross addition had been one of the major factors behind Grameenphone's revenue growth. However, this has slowed down recently. In the face of stiffer competition and declining voice ARPU, Grameenphone saw its existing subscriber base as one of its main competitive advantages, alongside network quality and coverage.
The company therefore decided to focus on maintaining and improving the profitability of its valuable subscribers by expanding the consumption of nonvoice and value-added services.
"To do this, you have to start by understanding what kind of customers you have and what they want," Khan says. "If you retain their business by making the right offers, this is more effective than recruiting new customers."
The company started doing churn-prevention analysis and microcampaigns based on microsegments of the customer base.
"At first the process and methods used were outmoded and time-consuming. The results were not all that accurate," Khan says. "You can imagine the amount of data we have with 40 million customers."
Khan's team recognized the need for a software-based solution and put a plan to management. They looked at seven proposals from five vendors. "It was a tough choice, but SAS had the edge on analytical depth," Khan says.
Grameenphone needed quick wins, but it also wanted to adopt a long-term, structured orientation toward advanced analytics that would allow the company to predict the results of competitive initiatives.
The first project used SAS® Enterprise Miner™ and SAS Forecast Server to create segmentation and churn-prediction models.
"We identified 10 broad segments covering 27 million of our subscribers based on their usage patterns," Khan says. "Then we created further subsegments, enabling us to run meaningful campaigns on them."
Grameenphone selected three of the broad segments and developed separate churn-prediction models with specific strategic objectives.
"We judged that if we could extend life cycles among these customers, it would have a very positive impact on profitability," Khan says. "To do this, you need to be able to predict which customers are most likely to leave and develop offers and incentives to retain them."
High return on investment
Results from that initial project were highly encouraging.
"Not only did we prevent many customers from getting into a revenue-silent state, but we were also able to increase revenue per subscriber," Khan says.
"We shot out 62 campaigns to target groups averaging half a million subscribers," he adds. "The take-up rate was 20.3 percent, compared to 3-5 percent for previous campaigns. With incremental revenue per customer of BDT 20 (US 25 cents), our return on investment was significant."
Grameenphone has also initiated campaign management with SAS Marketing Automation.
"This will shorten the time to market with our new campaigns, which is crucially important because we only have a very short window of opportunity to contact customers whom we predict are going to leave," Khan says. "Once they are gone, there is just no way to reach them.
"SAS gives us a complete value chain from analysis through to execution," Khan concludes.
Copyright © SAS Institute Inc. All Rights Reserved.