Was the big bang big data plan simply too much information for CSPs to take?
There are no dumb pipes any more, neither in telecoms or the industries its infrastructures were likened to. In fact, there is an argument to be made that the thought processes of the communications service provider (CSP) are now muddled by too many information options, while there is massive pressure for them to start going over the top to take on their new digital competitors, writes Nick Booth.
When CSPs become as shape-shifting, self healing, hyper intelligent and omnipotent as they have the potential to be, they will have achieved this extraordinary transformation on a big data diet. However, they are a long way from maturity and there are considerable growing pains, with CSPs being socially awkward and not nearly as intelligent as they think they are.
Studying data analytics will provide the best foundation for healthy revenue development and avoiding any debilitating resource wasting conditions. This is still a relatively new discipline and the first question of data analytics should be about where you actually apply it, says Manish Singh, the customer delivery head at Tech Mahindra.
There are so many layers on which you can get data from, says Singh, so priorities can be identified if they are categorised under three general areas of activity: the customer offering, execution management and finding out what’s happening on the network. While, in theory, data from analytics could be fed into policy engines which can orchestrate a redefinition of the CSP’s infrastructure, in practice nobody in the industry has achieved the capacity for that standard of software definition of networks yet.
A better policy for now might be to match the individual customers with their actual needs, says Singh. The level of data analytics skills available now makes it possible for much more through mapping of customers with their usage patterns, such as the types of application they typically use, the capacity they need, and how they vary by time and place. This is achievable now but, as with all analytics, it is only possible to expose this information if you ask the right questions.
The biggest data analytics mistake that CSPs make, says Singh, is to fail to use data to make the customer’s life better. The likely dissatisfaction that customers feel over a dropped call or a lack of throughput could be automatically predicted and nipped in the bud with an apology and counter offer. This not only quells customer unrest it saves the CSP a fortune on call centre calls. If a second was wiped off the average support call of a CSP, it could save US$1million per year, according to big data analytics company, Mu Sigma.
By not using data analytics to be pro-active with customers, CSPs are not deploying it to its full effect, says Singh.
Get some tools, advises Astellia. According to its research 59% of CSPs quizzed have no access or tools to get the relevant customer data to take informed business decisions.
Astellia claims it achieved “remarkable results” when improving customer segmentation for Zain Bahrain by using customer usage analytics. In order to understand the main interests of its customers Astellia first analysed customers’ application usage patterns. This examined variables such as the number of sessions, session duration, bandwidth usage and geolocation.
“We could clearly distinguish different types of behaviour,” said Astelia’s communications manager Esther Duvall. This empowered Zain to revise its CRM-based segments and create a totally new customer segmentation regime. The CSP now uses segmentation to recommend more fitting offers to subscribers, which in turn helps it to optimise its network resources and enhance profitability of data services.
Segmentation is a perpetual process, says Duvall, so clients must continue to fine-tune the different segments in order to optimise revenues and control the impact on the network. Data analytics helped Zain Bahrain to increase offer prices by 30%, boost sales by 17% and encourage 26% of current broadband customers to adopted a new offer. The bottom line was that the average revenue per user shot up by 34%.
Before you can ask the right questions, you have to make sure you have the right data, claims Jeremy Perlman, the vice president for Europe at Trifacta.
“Mobile operators struggle to use data analytics tactically as they have to deal with extremely large and diverse quantities of data,” says Perlman. A typical CSP might have a billion incoming call detail records each day and it takes data analysts a long time to clean and transform the data, so by the time a picture emerges from this intelligence the chance to act has gone. Trifacta specialises in getting diverse data standardised and claims it delivers it into the hands of business users in a tenth of the time. So a CSP can get vital data on app usage, networks and billing fast enough to take affirmative actions.
Yes, but which action do they take first?