According to McKinsey, companies that inject big data and analytics into their operations show productivity rates and profitability that are 5 to 6 percent higher than their peers. Although embedding advanced analytics into operational and decision processes has been proven to increase speed, agility and efficiency impact, many telcos are still some way from being able to use these technologies to their advantage.

The volume and complexity of customer data today require analytics methods that go beyond traditional reporting and querying to predicting customer intent and behaviour at the moment of engagement. Telcos must use customer analytics to optimize engagement throughout the customer life cycle.

Companies may not even know what information exists. Untapped data can lay hidden in various patterns, such as the use of social media and social content or in correlations between different data sets. A great deal of raw data is completely unstructured.

Michael Franzkowiak, founder and CEO of data scientists Contiamo warns that the large data pools accumulating within company’s risk becoming “a data cemetery” if not made suitable for operational use.

“The business potential of analytics is yet to be tapped fully by the telco industry,” asserts Sheryl Kingstone, research director, 451 Research says. “Telecom operators often require help to unlock the value of data, through targeted and actionable insights.”

This is where Telco Data Analytics & AI comes in. Big Data analytics provides an unparalleled opportunity to understand the quality and causes of customer experience. This parallel of conferences targeting the US and Europe markets will reveal how operators can unlock that value.

“Most, if not all, telcos have embarked on some form of ‘digital transformation’ journey in recent times; it’s the business challenge of our time,” says Analysys Mason principal analyst Justin van der Lande, a chair of the Europe event. “Telcos need to use data-driven advanced analytics as part of this process of automation, to answer fundamental questions about how to remain profitable in saturated and commoditised markets, and to make the leap to deliver digital-age customer experience, unlocking new growth from emerging technology.”

Leaders responsible for the development of network architecture from Nokia to EE, Vodafone and Turkcell will discuss the role data analytics plays in enhancing network efficiency and performance. Attendees can discover how to avoid network failure with predictive analytics and glean the latest best practice around network automation.

“The business potential of analytics is yet to be tapped fully by the telco industry”
- Rob McLaughlin, Head Of Digital Decisioning And Analytics At Sky

Operators know they need to use every network event, every customer behaviour, every signal to help their customers to do what they want to do - and stop spamming them with dozens of messages a month. The ‘data exhaust’ contains billions of individual actions which, when viewed together, tell us something about the people behind those actions.

“It’s tempting to simply aggregate as much of data as possible,” says Lakin. “But data has its own value chain; its own sources of supply and demand. A human-like profile is a rich set of structured data that describes a single person in holistic way, manufactured from the data exhaust created by that person’s actions.”

In other words, businesses can use their existing customer data to manufacture human-like profiles at scale. These profiles can drive unprecedented customer insight and personalisation. Because when customers are treated like individuals, they are much more likely to engage.

“It’s tempting to simply aggregate as much of data as possible”
- Lakin

Rob McLaughlin, head of digital decisioning and analytics at Sky, questions the concepts behind data driven customer insight.

Deep-dive analysis and insight presentations abound but how much of it results in improvements to customer experience? In the case that insight leads to a change in experience it is often on a macro level, perhaps fixing a process or improving a journey. Undoubtedly there is value here but there is an opportunity to operationalise the outcomes of the analysis that remain untapped.

Sky’s analysis of its customers illustrates diversity, not just in demographic, but depending on their context, mood, relationship, point in time.

“We find that our platforms and communications are insensitive to this reality; one size cannot fit all. Customers are diverse, vanilla is always wrong.”

He calls for data to be analysed closer to where the customer experience happens “If having knowledge of the customer is the starting point of an interaction, the rules of experience change,” he says.

McLaughlin explains that the objective is to use big data, artificial intelligence and machine learning, to automate how Sky makes product recommendations for its customers. With the right approach, customers are happy and profits at the firm increase. These two outcomes can only be achieved if Sky makes effective use of the insight it generates.

“In the instance that we’re lucky enough the customer wants to share their data with us, we have the opportunity to understand them and think about what they need,” he says. “That doesn’t mean that we pre-determine the products they must have – it means we simply present them with content and ideas, and we’re constantly looking to increase the relevance of those offers. I refer to this approach as maximising serendipity – we’re looking to increase the chance that the things our customers see are going to be the things that they’re interested in.”

“We find that our platforms and communications are insensitive to this reality; one size cannot fit all. Customers are diverse, vanilla is always wrong.”

Network problems are one of the chief causes of churn. Customer frustrations can ripple into customer support and network engineering teams. One of the most common is the implementation of VoLTE – or any voice service over an IP infrastructure – because of the greater susceptibility to problems and delays over the entire network (radio network, transmission, core, and application services) than the mature traditional cellular voice services.

As LTE radio networks and VoLTE services become increasingly critical to business, consultants Ovum recommend that operators “take proactive steps by investing in tools that will assure the QoE of these services. Being able to manage network performance in real time enables [operators] to deliver better experience to customers.”

T-Mobile US, for example, has deployed Ericsson solutions across its network of more than 72 million subscribers to measure end-to-end service quality in real time.

“Operators want near real-time insight into the quality of experience (QoE) that subscribers receive,” underlines Nokia. Its Mobility Analysis and Optimization tools are claimed to reduce dropped calls by up to 35 percent using crowd-sourced performance measurements and algorithms developed by Nokia Bell Labs. Furthermore, the modelling is used to quantify the link between video QoE and underlying business drivers such as churn, and NPS.

More agile and advanced methods are required to address the requirements of increasingly dense, complex, and multi-layered networks on the path to 5G.

Tele2 Netherlands, for example, is using Nokia Traffica's geolocation capabilities to provide a granular view of services issues down to the geo-tile level. This also provide a map-based visualisation system making it easier for service providers to precisely pinpoint the location of service issues, such as poor cell coverage and dropped calls.

According to Meile de Haan, CTO of Tele2 Netherlands, these capabilities mean the telco can better detect, troubleshoot and resolve network and subscriber issues using real-time data, “dramatically improving overall network quality, delivering a better subscriber experience and reducing churn.”

Network outages will cost money and reputation so identifying the root cause of the issue is a priority. Fortunately, massively-parallel GPUs - once only used for graphics processing - are capable of processing data extremely fast, and with extreme efficiency. Isreali telco Cellcom, for example, deployed the GPU accelerated monitoring solution of Sqream to deep dive into its network, diagnose and help resolve high incidents of call drop rates. Cellcom network engineers can now do this routinely and remotely, without having to visit the actual base stations to view logs.

In Singapore, Singtel’s data analytics subsidiary DataSpark worked with Analysys Mason to devise an app that helps operators allocate capacity by analysing datasets including customer satisfaction scores, potential demand as well as network traffic and yield.

Operators can choose the best location for new cellular sites or upgrade existing network infrastructure using insights gleaned from the data.

Ying Shao Wei, COO of DataSpark claims it’s a “game-changer” for telcos: “By analyse datasets, including past customer transactions in a non-invasive manner, operators can customise products and services for specific customer segments to generate new revenue streams.”

Telcos routinely perform forensics on dropped calls and poor sound quality, but call detail records (CDR) flow in at a rate of millions per second. This high volume makes pattern recognition and root cause analysis difficult, and often those need to happen in real-time, with a customer waiting for answers. Delay causes attrition and harms servicing margins.

Data management platforms from developers like Hortonworks can perform the heavy lifting for operators by ingesting millions of CDRs per second and processing them in real-time to identify troubling patterns.

Our mobiles know us nearly as well as we know them. The same data that connects mobiles to networks holds a great deal of value, and can help city planners create more liveable spaces, transport agencies provision better services and advertisers reach more relevant audiences.

This isn’t individual data, but rather anonymised network data. When the data from millions of subscribers is combined and any identifying features stripped out (phone numbers, for example), powerful insights can be unlocked into how the population behaves. Operators and city planners can create profiles for different segments of the population (young workers, retirees, etc.) for understanding travel patterns, optimising networks and providing updates to travellers, to identifying where best to open a new shop, based on potential customer density.

Applied to cities, this data can inform everything from smart tolls (based on traffic congestion) to urban planning and the provision of public services.

Deutsche Telekom has partnered with German city Bonn to provide Bonn residents access to data about the city. Specifically, the city authorities are using DT’s Data Intelligence Hub as a platform to collate citizen data sets and in turn provide local information about sight-seeing opportunities, the location of Wi-Fi hotspots and taxi stands and waste collection times. Lots of cities already collect a wide variety of data. They deploy traffic and environmental sensors to provide information on the traffic conditions and to collect data on air and water quality. Other possible applications include smart parking, smart street lighting and optimised waste collection.

DT subsidiary, the IT services and consulting company T-Systems will soon be forecasting train arrival and departure times for Deutsche Bahn trains. A learning algorithm will provide real-time forecasts for the more than two million stops made daily throughout the railway network, along with updated forecasts of available connections. Via a smartphone app, as well as via monitors at rail stations, Deutsche Bahn customers will be provided with real-time train-schedule updates up to 90 minutes in advance.

Nokia is among those pioneering new mobility analytics use cases to help operators create value from the data in their networks to address the needs of digital cities. It worked with Singapore-based operator StarHub to trial use cases, integrating into the Nokia AVA cognitive services platform and says it’s ready to customize the solution for other operators.

By using subscriber location information, operators are able to create value for digital city players such as city planners, transportation authorities and the travel industry. For example, measuring commuting patterns enables authorities to plan more efficient roads and public transport networks.

Business owners benefit from more effective advertising, targeting people in the vicinity based on their travel patterns, activities and personal interests. City authorities can identify the movement patterns of people to help decide where to build parks, shopping malls, recreational buildings and offices.

Tokyo-based NTT DOCOMO is the largest mobile service provider in Japan, serving more than 68 million customers. The organisation’s solutions include web service systems such as a voice-activated agent service called ‘Shabette Concierge,’ and ‘mydaiz’, business systems such as data analysis, and mission-critical corporate applications.

One of the company’s primary solutions is its internal data analysis system, which is used to collect data from a variety of internal and external sources. NTT DOCOMO data scientists access the system to analyse this data to improve marketing operations.

Since 2012, the operator’s web service systems and data analytics platform have been hosted at Amazon Web Services (AWS) which currently runs a massive 270-8xl-node cluster with over 4,500 virtual CPUs and 30 terabytes of RAM on Amazon Redshift.

NTT DOCOMO delegates some statistical analysis to Amazon Relational Database Service (Amazon RDS), and some of the company’s data scientists access data in Amazon RDS.

“Analytical queries are ten times faster in Amazon Redshift than they were with our previous data warehouse,” says Yuki Moritani*, manager of Innovation Management Department, NTT DOCOMO. “We have several petabytes of data and use a massive Redshift cluster. Our data science team can get to the data faster and then analyse that data to find new ways to reduce costs, market products, and enable new business. In our on premises environment, extending the data analysis platform to add new data sources might have taken several months to complete. We can now do it in weeks using AWS and Amazon Redshift.”

The use of big data analytics runs throughout the organisation. Other initiatives include the launch, in August, of a proof-of-concept (PoC) aimed at realising a new video IoT solution that will enable the interpretation and analysis of video data sourced from surveillance cameras using edge computing.

This data analysis is expected to be deployed in a wide range of scenarios, such as in general security, quality inspections during manufacturing processes, human presence detection, and marketing initiatives at retail stores.

Until now, the transfer and processing of large volumes of video data to the cloud was a lengthy process involving significant delays and placing a considerable burden on cloud infrastructures and communication networks. The new solution, according to NTT DOCOMO, will aim to address these shortcomings and herald a new era of high-speed image recognition and processing accelerated by edge computing.

The operator has also developed population flow statistics based on data from the company's mobile network to analyse how people move about on a constant 24/7 basis throughout Japan.

From this dataset, it became possible to estimate population flows in given areas, as well as specific routes, distances and speeds travelled. The expectation is that the statistics will be used for highly effective and precise planning.

For example, NTT DOCOMO has developed a system for taxi demand forecast in real-time which predicts future taxi pick-up demand in each area. It reduces the imbalanced operations between the amount of taxi supplies and trip demands.

The statistics could also play an invaluable role in planning for the expected increase in visitors to Tokyo in 2020 when the city hosts the Olympic Games. There will be a further surge in connectivity and IoT applications as NTT DOCOMO rollouts a commercial 5G network from 2020.

Learn more about NTT DOCOMO’s plans from Ai Hayakawa, Data Scientist, Big Data Group who is a guest speaker at Telco Data Analytics & AI USA.

*Yuki Moritani is currently working CSIS as a Visiting Fellow.

“Analytical queries are ten times faster in Amazon Redshift than they were with our previous data warehouse”
– Yuki Moritani, Manager of Innovation Management Department, NTT DOCOMO