The time for Artificial Intelligence is now
Artificial Intelligence (AI) is a hot topic these days, a very hot topic. Rightly so, as AI has the promise of being able to drive real innovation and optimization for enterprises great and small globally. Microsoft is setting its sights even broader with AI, where we are driving a strategy to make AI capabilities available to not just enterprise organizations, but to every business and human on the planet. Microsoft firmly believes that with the right source data, AI can be used to solve almost any problem out there, even those that affect the quality and future of life, as we know it here on earth. Therefore, with AI being democratized in this way, it is no real surprise that AI is enjoying such a lot of attention currently.
As this is a Telco & Media blog, it then raises the question of how AI and its companion technology, Machine Learning (ML), will affect the industry. How can telcos leverage AI to drive new growth, innovation and optimization? What makes AI so exciting for the Telco & Media industry is the vast amounts of data that telcos can access. This characteristic of telcos has long been touted as a benefit, but it seems that with AI, telcos can finally cash into the benefits that generating and collecting such vast amounts of data potentially enable. Having made the claim that AI and ML hold significant potential for telcos, we will discuss a number of opportunities (amongst many) that telcos should be exploring, and touch on a few potential pitfalls that must be considered.
Predictive Maintenance and Optimization
Telcos build, operate and maintain huge mobile and fixed networks as well as the data centers that hold their core computing networks. This infrastructure is extremely expensive to build and costly to maintain, both from an equipment and human resource perspective. Any technology that can reduce these high costs is therefore potentially very beneficial. Employing Data Analytics and Machine Learning, telcos can improve network performance and avoid downtime by gaining insight from streaming telemetry data. Being able to optimize the network in real time, called Self Optimizing Networks (SON), or predict when a site is likely to go down, can drive savings in infrastructure capacity and prevent revenue losses, which ultimately result in improved ROI (Return on Investment) for the large capital investments telcos need to make in their networks and data centers.
Subscriber Churn Analysis
Acquiring customers in a highly competitive market is extremely expensive, therefore telcos need to do all they can to prevent churn and ensure longevity with their existing customer base. Subscriber churn happens for a variety of reasons, such as price, service issues, customer support issues or disputes, lack of features in current offerings, features offered by a competitive service, privacy concerns, and brand loyalty (or disloyalty). These issues could be singular, or more often multiple issues lead to a customer switch; and these issues could be major and long-term, or minor and short-term (for example, the customer is just having a bad day). Trying to manage churn through offering blanket discounts or freebies is overly simplistic and ineffective. Each of these situations needs to be identified and handled differently – the price-sensitive customer can be given a discount or special offer, while the one with a dispute may just need to be given time to be heard. Properly identifying and managing each situation requires a data powered solution that can collect customer experience and billing information, and cross-reference this with other internal and public sources such as location and social media. Telcos need to use Data Analytics and Machine Learning to identify each churn scenario pro-actively and manage it effectively before customers decide to leave and potentially never return.
“Each of these situations needs to be identified and handled differently – the price-sensitive customer can be given a discount or special offer, while the one with a dispute may just need to be given time to be heard.”
With Microsoft’s ability to bring advanced Machine Learning capabilities to power users and analysts, it has never been easier to create, deploy, manage, and extend sophisticated churn models. The Customer Churn and Real-time Analytics solution provides a fully functional, end-to-end telco churn model that can be easily connected to backend data sources, configured for your specific needs and market conditions, and deployed to the cloud – no servers to provision or software to install.
Content Recommendation Engines – Many telcos are investing in media and content, driving quad-play strategies. The success of such initiatives relies very heavily on knowing individual customer usage patterns and content preferences and serving the appropriate content to subscribers. Telcos can use customer viewing, purchasing and rating information to perform real-time customer profiling, enabling them to serve accurate recommendations to customers, based on data points across many dimensions. This data analytics approach involves grouping customers into profile cohorts (based on past purchase and/or preference information, along with demographic information like gender, age, and location) to provide real-time recommendations (i.e., “people like you watched X”). Along with specific customer recommendations, this data can be used for internal customer relationship management purposes, to identify a customer who may be likely to switch to another service soon, or linking this data to other systems to enable Next Best Action (NBA) type up/cross selling insights. An example of this is MEO —the “quadruple play” brand of Portugal Telecom. Microsoft’s Data Analytics and Machine Learning solutions enabled MEO to build customer loyalty by providing innovative, personalized experiences that maximized the value that customers got from the services they purchased, allowing MEO to increase their Video on Demand (VOD) consumption.
Intelligent Agents (Bots) – Whether we like it or not many customers still like using the old-fashioned means of calling the Call Center to sign up for new services or resolve service issues. This places a huge burden on telcos to staff and manage these Call Centers. In many cases it also results in less than amazing customer service – I mean, we have all had some really bad experiences like this. Well, imagine if a telco could enable its Call Center with bots, which would provide a consistent customer experience by always being up to date with the latest services and offers, always resolving similar queries using proven resolutions, etc. In the past, this has been a bit of a pipe dream, as bots simply did not have the capability to effectively engage with humans. With advances in AI technologies and industry specialization, these challenges have been overcome. For example, the Amdocs SmartBot, in partnership with Microsoft, can engage customers over multiple channels (Messenger, Kik, Skype, Slack, SMS, IVR, Amazon Alexa, etc.) and is pre-trained on telcos’ business processes and telecom-specific intents, for engaging in intuitive, personalized, and contextual conversations with customers. This radically shifts the customer contact experience, bringing consistency to customers and reducing costs for telcos.
Having looked at some of the amazing AI initiatives that telcos could drive as part of becoming digital, a look at some of the potential pitfalls is also required. AI and ML can be immensely powerful and beneficial, but they are not a silver bullet for every problem. Just like any IT project, an AI project needs to have clearly defined goals and measures for success. Taking large swathes of data, throwing ML at it and hoping for miracle insights is not going to yield any fruit. The business need must be clearly articulated, and if AI or ML is an appropriate solution, and there is sufficient data of sufficiently high quality to support the approach, then magic could indeed result. If not, then look for solutions elsewhere. Another potential pitfall is having access to the right skills. Global tech firms acquire much of the top data science skills, and building the right pool of skills in-house could be challenging and expensive for telcos. An alternative could be to find the right long-term partner who could provide the required skills and competence as required, such as Microsoft’s Enterprise Services team. Telcos need to decide on which approach would work in their market, and then build their AI strategy and initiatives around this.
Finally, an encouragement to be bold. Becoming more digital across the entire enterprise is daunting. For telcos it can be very daunting indeed as the technology, services, partner ecosystem and organization can be extremely complex. However, the potential impact of AI for telcos cannot be underestimated. So, be bold, look for great opportunities to apply AI and ML, find the right partner and then take a step towards this cutting-edge way of solving problems and driving new value.
