The Future of AI for CSPs

The Gartner report “Market Trends: Top Five Disruptions for CSPs, Worldwide, 2018-2023” lists technologies like open-source software, virtualization, wireless edge computing, and artificial intelligence as radically altering business relationships and competitive structures.

Gartner believes that many CSPs still struggle to become early adopters or fast followers of these technologies. Legacy mentality is making it challenging for leadership at CSPs to develop the bold vision and agility required in order to capture market opportunities proffered by these trends.

What is real, however, is that the accelerating pace of change demands ever more rapid decision-making process. And these decisions must be based on facts – much of which is buried in considerable volumes of complex business data the organization may already have – it is just that they do not know what it looks like, where it is stored, how to access it, and what to do with it even it hits them in the face.

Gartner raises the proverbial question: How can substantial volumes of data from structured and unstructured sources be interpreted? The solution is to find patterns and relationships beyond the data and analytics acts as a driving force for this.

Evolving Use of Analytics

Charlotte Patrick, research director, Gartner, said marketing has always been a starting place for analytics – and then, with the advent of Hadoop, “customer experience analytics” has been a big focus. “Our analysis of use cases shows that channels offer the most immediate opportunities. This is mostly centered around the provision of more personalized or proactive interactions between the CSP and their customer, requiring new predictive/prescriptive analytics.”

She cited the contact center as one of the higher opportunities for data/analytics to provide value. It will use context mining from voice and text, real-time and predictive analytics that identify the reasons for calls and their likely resolution, as well as deliver prescriptive actions that the agent could take based on data about historic case resolutions. Interaction assistance tools bring real-time decision support for agents.

“CSPs have data from their operational support systems about the network, devices and services being used by customers, and this can feed into these tools – along with others – that enable issue resolution,” suggested Patrick.

There will also be more prescriptive analytics used to optimize assistance – for example, Nokia Motive’s Dynamic Intelligent Workflows analyzes data from previous workflows, the network, customer premises equipment and trouble tickets to find the optimal remediation of issues when subscribers contact agents with billing, subscription and network service issues.

She further explained that management dashboards will ingest and report on data for the monitoring of agents (for example, performance, hard and soft skills, among others) and customer experience (feedback management tools ingesting data from surveys and other sources).

Beyond the customer engagement center, Patrick noted that a set of more operational use cases around areas such as call routing will become more prescriptive. For example, data from the network, which highlights real-time issues that may see a rise in calls, can be used to better contain the calls (diverting them to recorded messages, for example).

“Another channel offering good opportunities for new analytical solutions is digital. Predictive analytics will enable CSPs to improve personalization, next-best offer and content recommendations and advert targeting. The digital nature of the channel also offers new types of data that can be analyzed to improve customer insights – for example, text or graphics,” she elaborated.

Other areas of opportunity include retail, sales, field services, finance, and fraud management.

AI – Not Ready

Much hope is being pinned on artificial intelligence and the perceived promise it will bring to businesses. Gartner estimates AI can bring in as much as $1.2 trillion in 2018, growing to $3.9 trillion by 2022.

Patrick noted that with so much data at their disposal, CSPs have an opportunity to reap the benefits of such data stores. In her view, the top three drivers for AI adoption among CSPs include more revenue, reduced costs and better customer experience.

Early applications already under review or consideration include machine learningand rule-based automations both of which are finding application in areas like network management. AI itself is finding new in customer service and customer experience.-Telecom Asia