The second European Customer Day at the Gaîté Lyrique Center in Paris brought together more than 400 Customer Care professionals from all over Europe for a day-long exploration of what the future holds around the theme of “Enhanced Humans.” Arnaud de Lacoste, Chief Marketing and Ventures Officer of Sitel, kicked off proceedings by tackling Artificial Intelligence and the related challenges for the Customer Experience industry.
Do we really understand Artificial Intelligence? Or Machine Learning…or Deep Learning, Natural Language Processing, Chatbots and so on. “These buzzwords keep popping up when we talk about AI, but once we get into the heart of the matter, we realise that there is still a long way to go. We are still at the very early stage”, explains Arnaud de Lacoste.
The European Customer Day was all about sharing and providing tools for understanding the broad range of buzzwords, which can often be more confusing than enlightening.
Put simply, AI is an old topic. Alan Turing, famous for cracking the Enigma code during the War, was one of the first people to theorize AI and to formulate reflections around “what will happen when machines take control away from humans.”
For 40 years, nothing really happened in the domain of AI. That was the “winter of AI” and now, for the past couple of years, we’ve been talking about the “spring of AI”. This is primarily due to plunging storage costs, enhanced IT processing power, the boom in basic research and Deep Learning techniques and the volume of accessible data that have driven and democratised Digital Transformation. As Arnaud de Lacoste points out, “the 1985 Nintendo console had the same computing power as Apollo 8 in 1968, and the 2010 iPhone 4 had around the same as the CRAY computers of the 1980s”.
“What really got me going on the subject was a headline concerning our industry entitled The End of the Line that appeared around 10 months ago in a major media publication”, reveals Arnaud de Lacoste. “Because I was constantly hearing that the days of our industry were numbered, I felt that it was really important to focus on the following question: “replaced…ok, why not, but by whom and for what purpose exactly? Who will replace our customer advisors in the future?”
At present, there are two major obstacles to having automatons perform a big chunk of what we currently do. First, we should bear in mind that voice recognition software is not 100% reliable and the bulk of a customer advisor’s work requires nearly 100% accuracy, especially when we’re dealing with simple subjects (for example, nobody is going to tolerate 80% accuracy during appointment scheduling – a straightforward task. It has to be 100% accurate). It would, therefore, appear that at present, very few tasks can be readily automated. Moreover, Natural Language Processing (NLP) only processes about 7% of the elements that we use to communicate – voice and intonation account for 38% of our expression. Indeed, the intentions and emotion projected by the voice structure enhance the Customer Experience, so we are missing an important part of customer understanding. “I firmly believe that we have entered the era of complementarity and this means collaboration rather than substitution”, declares Arnaud de Lacoste.
“We hear that AI is set to revolutionise 30%, 40% or even 50% of our operational platform activity in the next 18 months and again, I think we’re making two mistakes here. The first is that we are looking at processes from a human standpoint, even though what may appear simple for a human could be very complicated for a robot and vice versa. We need to abandon our human perspective and look at things from a robot’s standpoint when we plan on automating certain tasks. The second error is to view work and time in a monolithic manner. Our work is not a monolith, but is composed of hundreds of tasks that cannot be 100% automated.”
To achieve the required 100% accuracy, emotion, situational understanding and empathy are all crucial in creating a balanced and appropriate Customer Experience. There are more than just processes at work. Describing the work of customer advisors as simply a series of tasks is a big mistake. The key to successful Customer Relations and Customer Experience now lies in this complementarity between man and machines.
Arnaud de Lacoste continues, “over the past few months, I have probably met with around 80 start-ups throughout the world that focus on AI and as yet, none of them have come up with a magic formula. And this is why we have devised the botshore concept – a new automated production space that combines both the human and the technical. The shore analogy – onshore, nearshore, offshore, etc. – is striking and represents a future that is not too far off.”
Several parallel processes will be used to process interactions: 100% automated, semi-automated or “bot-assisted human”. The latter may be used at front office level to deal with customer queries and at back office level to enable care advisers to work faster or to provide customers with more effective solutions. The bot will listen to the conversation and prompt the care advisor based on the customer profile analysis. With non-linear customer journeys, there will be constant toing and froing between these dimensions. The areas of man-machine cooperation will multiply and customer care will be underpinned by processes for the part of the conversation that will be initiated by the bot, with the customer advisor taking care of the rest. “We firmly believe that our care advisors are best placed to know whether the bot is being helpful or not, or whether the solution is the correct one. The bots will need to be trained, coached and supervised. And what better trainers than the expert advisors themselves! So we could envision a situation where the work of a customer care advisor would consist of “managing bots” – they would be responsible for ensuring that the bots which work alongside them are functioning properly. They would only intervene in order to resolve problems being encountered by their ‘bot team’.”
The Customer Experience industry is indeed faced with some major challenges, but many of these are positive and rich in future potential.
The first – which is already in progress – involves new interfaces. Arnaud de Lacoste jokes that “Andrew Yaroshevsky, one of the founders of Chatfuel, came up with a great quote to explain the craze for messaging over Apps: downloading an app is like a marriage, messaging is more like a date: just a one-off with no long-term commitment.”
So we will move from start screens loaded with different apps towards screens that display conversation threads with our peers, friends and with brands within different messaging platforms.
Interfaces will be increasingly digitised with fewer and fewer screens – take Google Now or Amazon Alexa for example – with no search button or ads. We will, therefore, need to devise new customer relations processes together with the corresponding journeys and trajectories. As Arnaud de Lacoste points out, “we don’t speak to a device like Echo in the same way as we do to a telephone or to a screen”.
The second crucial development relates to data and closely related notions of context and proactive relations.
By 2020 there will be 6 billion cell phones and 20 billion web-based devices generating a colossal quantity of data with four essential features: volume, variety, speed, and veracity. Lots of data in connected conversational interfaces means a multitude of propositions and possibilities. “Data management strategies need to be proactive. It will be all about coming in at the right time and actually being of service”, stresses Arnaud de Lacoste.
Finally, the third trend, the logical corollary of data processing, i.e., customer trust.
The “Trust Economy” that everyone is talking about lies at the heart of Customer Relations and is underpinned by three interconnected pillars – data privacy (including cyber security), transparency and reciprocal exchanges – that demonstrate the importance of brands’ commitment to respecting the values of their customers over their own processes.
These pillars impose new standards. Brands will not only have to secure our data, they will also have to reveal everything they know about us. Arnaud de Lacoste believes that “the winning brands will be those that actually show customers all of the data to which they have access and explain how they use the data for the customer’s benefit. It’s not the abundance of the data that counts, but making good, hyper-personalised use of it. In other words, ‘better data’”.
We need to beware of buzzwords when dealing with the challenges of Artificial Intelligence. “AI technology will very soon be accessible to everyone. But it is data that will confer it with value – your own data, data on your employees, your knowledge, and your know-how: an irreducible set of values that will continue to be more important than pure technology, which serves no purpose on its own. The digital/human tandem is the winning combination”, concludes Arnaud de Lacoste.
The future of the Customer Care Experience will be built around man-machine synergies: a vision that was shared and reiterated in the successive presentations given at European Customer Day 2017 by Facebook, Air France, Groupe Rousselet, Voyages-sncf, Edenred, Royal Air Maroc, iDTGV, Direct Assurance and others.