The top three customer service lessons from the Artificial Intelligence World Summit
• Experimentation, regulation and monitoring were the hot topics at the recent AI World Summit in Amsterdam
• Bas Hiemstra, manager in Arvato CRM’s Automated Interaction team, reports on the most important lessons to take from the Summit
Read time: 6 minutes
The brightest and best talent in the world of artificial intelligence descended on Amsterdam last month for the AI World Summit. More than 6,000 people from across blue chip businesses, tech start-ups and science came together from 161 countries to tackle the biggest questions and challenges in the AI space, and the packed two-day agenda included keynote speakers from Microsoft, NASA and Google.
We’ve previously talked about how AI is already a major force in changing how customer service looks. It’s clear from the World Summit that this trend is only going to continue.
Bas Hiemstra, portfolio manager in the Automated Interaction team at Arvato CRM, is dedicated to bringing the latest AI innovation to our customer service solutions. He went along to the Summit and here, he reports back on the most thought-provoking, useful and need-to-know take-outs for everyone in the customer service space.
1. Don’t just jump on the bandwagon. Start by experimenting with how AI fits into your customer service strategy.
“AI is a buzzword that seems to be on everyone’s lips and at the top of every forward-looking business’s agenda. That’s understandable, given the massive potential benefits that AI can bring. But, as Dr. Joseph Sirosh, Chief Technical Officer for Microsoft AI, said in his keynote speech, there’s a danger of jumping on the bandwagon for the sake of it.
“Instead, it’s important to consider where and how AI fits into your customer strategy.
“Implementing AI into your business can be a real investment in time and resource. Yet it takes more than hours and money to make sure it’s a success. Conversational AI – including chatbots and voicebots – requires a wide variety of human voices and inputs to ensure its ‘personality’ reflects the business’s values. Microsoft’s AI assistant Cortana was trained and developed by a team that included a poet, a novelist and a playwright.
“One of the biggest pitfalls facing businesses is the temptation to patch AI onto complex existing legacy systems as a new and standalone channel, rather than baking it into the very heart of your business. This can leave you doomed to fail.
“An ill-thought out AI system can also cause a real headache for customer services representatives. We know that having a single customer view is the most beneficial for customers, but having all information in one place also makes the job of a customer service representative much easier. Layers of extra systems that aren’t integrated properly can be extremely disruptive and make extra work for the representative that has to deal with the customer complaints they cause.
“It’s better to begin your AI journey through small-scale and simple experimentation. Understanding what AI can be best used for, then identifying the areas in your business where it can be most successfully deployed, such as handling simple and straightforward customer queries via a chatbot, will likely lead to a better outcome for all.”
2. Continually examine and refine the outcomes of your customer service AI system.
“We’ve mentioned above that setting up AI effectively can be significant. But to really reap the benefits the work starts once the system is established. After all, the use of AI is only as good as the outcomes it delivers. For example, does it actually effectively resolve simple customer issues and free up customer service representatives to handle more complex queries?
“There’s also the danger that, when AI systems and chatbots are typically used to handle straightforward questions, businesses don’t pay attention to their outcomes. They are actually a valuable source of insight about recurring issues your customers face. Continually reviewing and monitoring the queries a chatbot deals with means you can tackle common problems before they arise.
“And don’t forget that your AI system or chatbot might learn patterns and trends that have unintended outcomes. The recent news story of Amazon’s AI recruitment tool is a prime example of this. The tool was trained to review applicants by observing patterns in resumes submitted to the company over the last decade. The majority of the data it had to learn from was applications from men, simply because more men applied to work at Amazon than women – but the tool took this information and ‘learnt’ to favour male applicants.
“This example is an extreme one but shows how crucial it is to continually monitor how your AI systems are behaving in practice. This is especially true for customer service, where patterns learned by chatbots, for example, are on display through interacting with consumers and could cause genuine reputational damage.”
3. With great power comes great responsibility to keep the human voice in customer services.
“It’s a cliché, but it’s true – the potential for AI to revolutionise our lives brings both great promise and risks.
“The number of non-governmental organisations and political groups represented at the World Summit shows that the topic of AI is not just important for businesses and profit-making companies, but something that has enormous potential for the public sector and society at large. Accordingly, the overall theme of the summit was #AI4Good – making sure that AI is developed in a way that makes the world a better place
“It’s no surprise then that ‘responsibility’ was a recurring topic of conversation, with lots of debate about how the use of AI can be appropriately regulated without stifling the potential for businesses.
“California recently passed a bill outlining the guiding values for developing AI in a safe and beneficial way. It has also mandated that bots have to clearly disclose that they are not a person when they are interacting with customers.
“We’re unlikely to see this level of regulation across the board, and in reality, businesses will probably have to self-govern their use of AI.
“There’s still some public nervousness about ‘robots’ taking the place of humans, particularly in the field of customer service, so businesses are treading a fine line between making best use of the technology while not alienating their customers.
“The European AI community is already facing up to this challenge with the recently-launched CLAIRE initiative. It champions the development of artificial intelligence that complements, rather than replaces, human intelligence.
“There’s a clear commercial imperative to getting this right. Customers still want to hear a human voice when they interact with a company, and the businesses that succeed will be those that manage to maintain the careful balance of harnessing technology to be more efficient with the warmth and personality of the human voice.”
Author: Editorial team Future. Customer.