At the Hadoop Summit Conference this week, AT&T Vice President of Big Data Victor Nilson outlined the ways in which AT&T are using Big Data to cut costs and improve customer experience.
The volume of data in the AT&T network has skyrocketed by over 50,000% in the past six years. In the same time frame, they have invested $40 billion in their wireless and wireline networks. The company collects 30 billion data points an hour, and uses this data to work towards optimising efficiency and customer experience.
One way in which this data is put to work is in a tower-outage analyser that AT&T have developed. Traditionally, gathering information on how tower outage impact customer experience has been challenging, as calls are often simply added off to nearby towers. The tower-outage analyser offers a data-driven picture into customer activity in the affected area. Using this technology, AT&T can prioritise repairs in the areas with greatest customer impact. AT&T claim implementing the tower outage analyser has led to a 59% improvement i customer experience.
Another area in which AT&T are harvesting their big data insights is motor vehicle repair. AT&T possess ten of thousands of service vehicles, and the cost of keeping them on the road is pretty hefty. They spent $7 million on towing charges and $10 million on battery jump-starts in the last year alone. So, AT&T have developed a system which collects data on 80 different metrics, to proactively discover which vehicles need maintenance, and not kind of maintenance they require. They can now, for instance, replace batteries before they die, cutting down on the jump-start bill. Nilson describes this process as “moving from time interval maintenance to just-in-time maintenance”.