So, what\u2019s the reality here?<\/p>

If you talk to CSPs there\u2019s an honest answer. Indeed, when this finding was shared with a CSP at an industry forum, the answer was: \u201cWe\u2019ll take the plaudit but there\u2019s a lot more that we\u2019ve got to do on AI\u201d. Indeed, in that same study, most CSPs admitted they felt they are lagging behind their competitors in their use of AI and are playing catch up.<\/p>

CSPs seem to be enjoying a halo effect from how their connectivity and content services are associated by consumers with the more sophisticated digital assistants like Amazon Alexa and Apple Siri offered as part of devices that they heavily advertise.<\/p>

This halo effect is a mixed blessing, however. As AI becomes more integral to all kinds of consumer devices and scenarios as wide ranging as connected cars, some with autonomous driving functions which even interact with traffic lights, or smart in-home AI assistants that control your home as well as listen and respond to your commands, there will be a steep up-levelling of consumer expectations from AI. AI is becoming the new normal for customer engagement and the pressure is on for CSPs to match these expectations.<\/p>

The challenge is that getting AI right for a CSP isn\u2019t easy but clearly the industry needs to respond. So, where to start? <\/strong><\/p>

The obvious answer may seem to be where customers experience it for themselves. In other words, those customer-facing interactions that can be done by software robots or chatbots and are analogous to how consumers are interacting with AI consumer devices in the home.<\/p>

While this is true, AI development and deployment needs to be progressed along twin but interlinked paths, focusing, on the one hand, on augmenting the AI experience for consumers by offering more humanlike bot services, and on the other, on augmenting the experience for human agents in care, sales and marketing roles.<\/p>

Human interaction is valued by consumers. This is supported by our study that revealed that while consumers find bot service to offer more convenience and speed, they are underwhelmed with today\u2019s bots and, if given the choice, prefer to interact with human agents.<\/p>

The experience is a mixed one also for agents. Part of the issue is how long it can take to train agents to use the systems and data at their fingertips to better serve customers. It is estimated this can take up to nine months. However, with staff churn rates averaging at six months, customers may rarely interact with a fully trained agent, explaining how the experience can be mixed for both the customer and agent.<\/p>

This is where AI can step in and be the assistant to the agent, riding alongside on the call or chat session to make next-best-action and other recommendations in real time. AI can see, gather and analyse data much faster than a human ever could. So, by sitting on top of the data streams and proactively engaging with the agent, it can deliver an augmented experience for the agent. It is like how Tony Stark can don his Iron Man suit and become a super hero with augmented powers.<\/p>

Perhaps Marvel\u2019s Iron Man is too strong an analogy but it is true that AI can help to lift off the burden of repetitive tasks and, more importantly, significantly simplify complex ones and allow the business to do a much better job at using data to improve service personalization.<\/p>

Personalisation is key to how CSPs can increase customer satisfaction and retention and improve cross- and up-selling. To date, next-best-action functionality offer agents a relatively narrow set of recommendations. AI could really change this, doing the super heavy-lifting on the increasingly larger volumes of real-time data available that could never be properly analysed by a human fast enough to offer agents a much more differentiated set of recommendations, personalised to the end customer.<\/p>

For a CSP today this proliferation of AI, used on the inside by agents and interacting with customers at the front-edge of their operations may still seem fanciful, especially against what Apple and others are able to do with it. It is important to reiterate that AI is an iterative approach because it is about a machine learning. <\/em>Each new development or service may be one step forward or even one step backwards. But, this is the process by which AI self-develops and becomes smarter, stronger and more relevant. The important thing to do, is start on this journey sooner rather than later with the use cases that deliver real value.<\/p>

Projecting five years into the future, where will AI have taken the CSP by 2022?<\/strong><\/p>

Fundamentally, CSPs should have used AI by then to nail down self-service, offering customers much better self-service than today. When a customer asks a question or reports a problem, AI will pretty much fix it. And if it can\u2019t, it will do a seamless, clean handover to a human agent. It is unlikely that the need for human judgement, problem solving, and intuition is going to end altogether. This talent and the ability of humans to empathise needs to be tapped, developed and augmented through AI assistance.<\/p>

Given how AI could crack the personalisation challenge, AI-infused customer service systems could be delivering what might be termed hyper-personalisation. All the possible contextual data that\u2019s available and feeding off data streams from a universe of sensors and sources could mean personalisation takes account of the richest details affecting a customer at that exact time and place.<\/p>

Cutting-edge AI around fields like deep learning is already entering the mainstream. The capability of even smarter machine-learning must be considered. For example, developments in exotic fields like generative adversarial networks that use neural networks to create accurate copies of complex artworks could have a role in AI for CSPs. This technology is able to learn and differentiate extremely unstructured data, in this case brush strokes and pigments. Think about how a bot could use this same deep learning to super-personalise customer interactions. For example, it could instantly detect someone\u2019s dialect or accent and replicate it, speaking with that specific customer like someone from the Bronx, Merseyside or the posher parts of Paris!<\/p>

The next few years are going to be critical for the industry. In markets that are saturated and where CSPs need to develop multi-play offerings that are more complex to run, how AI can simplify and enhance the way the business engages with its customers and optimises workforce and processes could be the difference in winning or losing the business overall.<\/p>","blog_img":"","posted_date":"2018-01-19 12:08:35","modified_date":"2018-03-29 16:27:45","featured":"0","status":"Y","seo_title":"How Augmentative Intelligence staffs telcos with super human agents","seo_url":"how-augmentative-intelligence-staffs-telcos-with-super-human-agents","url":"\/\/www.iser-br.com\/tele-talk\/how-augmentative-intelligence-staffs-telcos-with-super-human-agents\/2833","url_seo":"how-augmentative-intelligence-staffs-telcos-with-super-human-agents"},img_object:["","retail_files/author_1516343478_temp.jpg"],fromNewsletter:"",newsletterDate:"",ajaxParams:{action:"get_more_blogs"},pageTrackingKey:"Blog",author_list:"Pankaj Lamba",complete_cat_name:"Blogs"});" data-jsinvoker_init="_override_history_url = "//www.iser-br.com/tele-talk/how-augmentative-intelligence-staffs-telcos-with-super-human-agents/2833";">