There is a huge expectation among operators in emerging markets that smartphone penetration will increase fairly rapidly over the next three to five years. Operators want to be at the forefront of this and are therefore investing heavily to ensure that they not only have the best networks possible, but also have a clear understanding of mobile data services in order to develop new business models to accommodate their anticipated growth.
There is a huge amount of traction for mobile broadband services in Africa and elsewhere in emerging markets. There are few fixed-line networks so mobile networks are very prevalent as a means of broadband access; many emerging markets are investing hugely in mobile networks, with mobile data in particular seen as strategic in terms of future growth.
Operators are aware that they need to invest in new technology but may be unconvinced that they can actually make any money by doing so. This is where Big Data Analytics comes in – it allows operators to ascertain ways of driving efficiency across their business while keeping their operations sustainable.
“With LTE networks, driving revenues is extremely important – having a high-quality network is meaningless unless there are good business models in place to drive profitability”, says Oliver Finn of customer experience specialist The Now Factory. “In a lot of emerging markets, LTE is just finding its feet, but operators are looking at mobile data as a means of differentiation as well as securing their profitability going forward. Increasing revenues means increasing usage, and ultimately this involves increasing the value proposition for customers.”
Particularly in emerging markets, customers have typically been treated as a number; individuals have until now not generated enough revenue for operators to worry too much about customer retention. However, the growing prevalence of data means that it is now in the operator’s interest to encourage its customers to use its network as much as possible.
Churn is still an issue – after all, it’s difficult to encourage use among customers who aren’t on your network. Therefore, while operators obviously want to retain existing customers they are also looking to encourage new customers onto their networks. In emerging markets, there is little operator loyalty, so in terms of churn prediction data analysis can be very useful.
By using available data in more imaginative ways, it’s easier to predict which customers are going to churn. For example, it would be possible to look up the call centre numbers for rival networks and check the subscriber base to see which customers had investigated changing networks, or see which customers had called the network’s own helplines.
Steven Totman, of Big Data Analytics firm Syncsort, observes: “For many firms, the term ‘Big Data’ refers to the thorough analysis of data that previous was essentially ignored. Data has been growing for some time, but the available tools have struggled to handle this.”
This data can now be used to the operator’s advantage – churn prevention being a major example. The more an operator knows about how a customer uses their network, the easier it is to provide them with an incentive not only to stay on the network, but to increase their usage too. In terms of core infrastructure, it doesn’t really matter to a provider whether a subscriber is making a voice or a data call – they care about active users and ARPU.
“Queen Bee analysis is particularly interesting”, says Totman. “It refers to users in the subscriber base influencing activity around them, in the way that a queen bee controls the workers around her. You could have a fourteen-year-old on a low tariff who will likely only use her phone for texting, but if she were to move to another network then she’d likely bring a lot of friends with her – young people in particular can have a strong influence over their friends, as they all want to stay in touch.”
Spotting the ‘queen bees’ – i.e. the users who communicate with and influence the most people – allows operators to target them specifically. By identifying ‘high-influence’ users and in turn influencing them – for example with a free upgrade – operators are able to increase the user’s loyalty. This is also likely to have a knock-on effect – users influenced by a ‘queen bee’ may seek to upgrade their handsets as well.
In emerging markets, data usage is only increasing and there are huge volumes of traffic coming onto networks. This is happening in the context of a very competitive environment; voice and messaging revenues are typically either flat-lining or decreasing thanks to OTT services, and profitability margins are very tight.
The huge volume of traffic is putting pressure on existing infrastructure which is forcing the migration towards LTE networks. Infrastructure used to be a barrier to new players entering the market; this ceased to be the case in developed markets some years ago and now the same thing is happening in emerging markets.
“Not owning infrastructure is no longer an issue – data has become a more powerful asset”, says Syncsort’s Nejde Manuelian. He notes that traditionally, legacy vendors have used data warehouses to store and analyse data. New vendors meanwhile are often starting with a blank canvas, which can be an advantage.
“They are not hamstrung by old architecture, which can be difficult to bring into the new era”, says Manuelian. “This grants more agility and flexibility – it’s very easy to put together a data warehouse that acts as the core of an organisation’s data value.”
By using data in inventive ways, it’s clear that operators can break with tradition and explore new business models. The value that data represents to an operator is clear – innovative analytics can maximise the revenue from individual customers as well as improving the overall customer experience.
Manuelian notes that “today’s buzzword is Big Data” – it’s not hard to see why. As LTE networks come on line and increase mobile network data capacity in emerging markets operators need to find ways to increase customer loyalty and reduce churn. Big Data would appear to be a key part of the solution.