• December 17, 2015

The Telco Challenge II: Instant Responsiveness to Customer Needs

Big Data and Privacy

As more of the conventional services move to Over the Top (OTT) or IP-based delivery, operators need to know in greater detail about the sort of traffic or applications their subscribers use.

HPE Telco Big Data technology extracts this information from the network and correlates it with more static information about the profile of the individual subscriber gleaned from interactions across multiple channels including customer call centers, retail stores and social media. With full insight about the experience and the usage habits, providers can decide, per individual subscriber, what the best course of action is to give a customer what they need.

The customer service representatives can choose — depending on the profile of the subscriber — to upsell, or to instantaneously resolve a situation by execution of a network policy, or even escalate the issue to the Network Operation Center (NOC) to handle as a high priority. Factors such as what the customer has recently posted on social media, how alike they are to other customers who have churned to take their business elsewhere and even the stress levels present in theirs and a customer call center operator’s voices — to whom they are making a complaint — can all be taken into consideration. This helps the customers get their problems resolved sooner with greater understanding by the telco, which also benefits from lower services costs and greater customer loyalty.

Another example of where this information can be used is in segmentation, whereby customers are aggregated into a small number of segments with other customers who share similar characteristics.

This allows telcos to target specific promotional messages for the latest phone upgrades or new service plans, with a higher degree of certainty that the offer will be accepted — thereby maximizing the number of services a consumer has with a telco at minimal cost. Customers get what they need just as their need emerges, and the telco gains access to additional revenues and ensures that their customers stay happy, thus benefitting both parties.

Although analytics is a very powerful tool, some telcos are understandably apprehensive about making full use of it, for fear of violating their customers’ privacy.

“One of the tricky things — and there’s a debate going around in a lot of telcos about this is staying on the right side of the ‘creepy’ line. It’s one thing for a telco to know you well enough to make you a valuable offer just when you want it; that’s useful and helpful. It’s another thing if that same information is used to a person’s disadvantage. Consider: if you just know where a person makes his calls, where their phone is, and the numbers he calls, you don’t even need to know the name associated with that number — you can infer a lot of information: where he works, where he lives, whether he exercises or goes to the gym, even things like if he’s having an affair because he’s spending time at someone else’s house,” remarks Dr. Kermode.

For a telco, it thus becomes very important to tailor its offerings and communications with
customers in a way that creates value but doesn’t cross the creepy line.

Branding and Consistently Excellent Service

In order to satisfy their customers, telcos must have a consistent, effective branding and marketing strategy. From a marketing perspective, this comes down to being able to integrate brand perception and the feelings about the brand. Thus, TV advertising, billboards, and other major visible marketing channels have to be closely integrated in the way they present the telco’s offers.

Also necessary is absolute consistency in the way the call center staff interact with customers when they call in, either to upgrade or to complain. Understanding every interaction with the customer over the past several months and seeing the pattern to spot issues precisely is critical.

“You should know if they’re stressed. There’s got to be culturally specific and sensitive handling of people when they call in. And you’ve also got to have that same consistency happening in the retail stores, because some people don’t want to talk on the phone, they want to go and talk to someone in the retail store.”
—Dr. Roger Kermode, Chief Technologist and Strategist, Communications, Media & Entertainment (South Pacific), Hewlett Packard Enterprise.

“We’re even starting to see the deployment of data and dashboards around these sorts of experiences, into not just the call center, but actually into the retail space, where retail staff run these applications on tablets and smartphones as they assist customers. People expect to be served well and if they’re not, they will take their business elsewhere,” he suggests. “So, yes the overall customer experience is definitely significant and telcos need to deliver a consistently good experience at every touch point.”

Success Stories

There are definitely cases of telcos that have been able to adapt to the new way of handling
customers, and among them is that of Rogers Telecom of Canada.

Rogers’s aim was to extend the reach of their data plans to subscribers that did not have them. For subscribers, avoiding usage of data is usually either because there are concerns about unclear tariffs, or because they simply haven’t had it enabled.

Whenever a subscriber tries to access Facebook, for instance, the HPE Subscriber, Network & Application Policy (SNAP) software solution recognizes that the customer is not subscribed to a data plan and instantly prompts the user with an offer like: “Facebook data pass for today? It’s only 50 cents.”

Being faced with a good value right when needed makes it an easy “yes” for the customer. Once approved, the day pass solution immediately applies both a network policy and a real-time charging rule to enable it.

Rogers implemented this solution with impressive results: it saw a 26% growth in unique users of data-passes, a 17% growth in data-pass revenue — despite a 30% decrease in price — and a 10% conversion to recurring monthly plans. It also saw a 98% reduction in credits for unexpected charges (as the tariff plan was simple and there was no room for misunderstanding).

Dr. Kermode explains further: “They also have a system that’s able to detect when people come into the country — from either their own customer base via a related operator or from other networks. So let’s say you’re travelling to Canada from another country. The system enables you to look at the amount of data that people have used, or the fact that they are from another environment, such as another carrier.”

“Say you’re from Singtel. You’ve gone from Singapore over to Vancouver. When you show up there, it would offer you a local data plan as opposed to an international one. They’re able to take quite a lot of cost out of international data roaming charges and for people who are on prepaid plans or locally, it’s able to proactively offer you new plans based on your past consumption over the last month or two. The goodwill and loyalty from these programs is very significant.”
—Dr. Roger Kermode, Chief Technologist and Strategist, Communications, Media & Entertainment (South Pacific), Hewlett Packard Enterprise

Another successful example of understanding customer experience, according to Dr. Kermode, is that of Orange Telecom in France. Orange used a system based on some of HPE’s analytics offerings that is able to look at what people are saying to the customer call center agents and to determine whether they are stressed or whether the call center agent is stressed. This integrates all their past experiences with the customer across the network, whether online, with the person in the store, or a previous call center person.

“With this system, you can look across and see whether the customer is stressed now as opposed to before. Also, you’re able to see if there’s a performance issue with the call center person — whether they’re stressed and how they’re reacting to the situation. And this has speech-to-text behind it as well. So very, very subtle sets of data points can be merged together to work out whether someone is going to leave or not,” he concludes.