Artificial Intelligence (AI) will replace customer support agents.
How often have you heard this over the last few years?
That’s usually the first thing people think when you mention AI in customer support. It’s as if machines were on the verge of completely taking over every single task and deal with all your customers directly without any human interference.
Yet, as someone who has spent the last few years developing this technology, I believe that this pessimistic vision belies the true impact AI will have on customer service. Not only because AI is still in its infancy (the AI revolution hasn’t happened yet), but because it also misses the main reason for implementing this technology: employee experience.
Until quite recently, all customer support managers cared about when looking at AI was customer experience. It would reduce costs, make things more efficient, increase revenue, and help build brand loyalty. But what about your team? What will happen to customer support agents who have been tirelessly working as hard as they can to fix your customers’ problems?
Well, according to Forbes’ 10 Customer Service predictions for 2019, “employee experience will be as important, if not more so, than customer experience.” And we all know how difficult the job of a customer support agent can be. They are under a constant crossfire, often dealing with impatient customers and complex problems. So if their job is to keep customers happy, they better be happy about their job.
This is where a significant potential of AI in customer service lays. If you want to keep your customer support team motivated, happy, and productive you’ll have to look at how Artificial Intelligence can make their work less stressful and more productive.
And this is already happening. During the last couple of years, we’ve seen many significant developments in this field which have helped many customer support agents become better at their job. From predicting the customer’s behavior to prioritizing tasks to chatbots and multilingual customer support, here are some of the most innovative technological developments that are about to shape the future of customer service.
Managing your customer’s anger
Nobody likes to deal with angry customers, and while it isn’t necessarily rocket science, it also presents a unique set of challenges. You have to weigh your words carefully, set the right expectations, and often go the extra mile to fix the issue.
This is something Artificial Intelligence can help you with. By using sentiment analysis, a field within artificial intelligence that uses algorithms to determine if the tone of a written message is positive, negative or neutral, along with machine learning, you can put the best-suited agent on the most urgent cases in a matter of seconds.
But how does this work?
First, sentiment analysis algorithms are able to sort tickets by urgency based on the email’s tone of voice. This way, agents can turn their attention to the most frustrated or dissatisfied customers without having to manually assess their priority. Second, by using machine learning, you can analyze all the interactions between agents and customers and identify patterns for the most successful ways of dealing with some of those issues. The AI system can, therefore, help you lead the discussion in the right direction, with the right tone of voice, by recommending a specific reply.
This will enable your team to deal with urgent matters faster while maintaining their sanity.
Getting your priorities straight
Customer support is a tough job and it can be even harder when you have a formidable volume of customer support tickets, phone calls, and live chat messages coming your way. Keeping the same response times and level of customer satisfaction, under these circumstances, is definitely challenging for customer support agents.
Good thing that machines can now give a helping hand during the busiest periods. By predicting the urgency of the requests and prioritizing accordingly, you can turn a once stressful process into a far smoother one.
The urgency of customer queries can be tackled as an automatic classification problem based on the customer’s historical data, such as recent transactions, and variables unique to each business, such as a product being out of stock. For example, if your business is a short-term accommodation platform and your customer’s booking was denied by the host, on the day before his or her arrival, your automatic classification system will move this up on the priority list and automatically recommend a few alternatives.
This will help customer support agents deal with multiple issues at the same time and never let them miss an urgent query.
Matching agents with customers
No two support agents are the same. Some might be good at persuading customers to buy more, others are better at fixing technical issues. But regardless of how great your team is, you’re going to want to match your agents’ skills with your customers’ needs to increase the chances of an issue being fixed as quickly as possible.
And this is something which can be addressed by Artificial Intelligence. First, you create models for each type of agent by mapping their behavior. Second, you analyze your agents’ behavior and customer satisfaction, to predict which model will better fit a specific customer request. Third, you optimize your current prioritization scheme by taking both models into consideration.
In the future, AI will definitely help optimize your team for better results, and this is just another example.
The bot revolution
Chatbots were meant to be the next big thing in customer service. They were the first AI technology to be massively implemented in this field.
However, at the time, most companies simply rushed towards implementing this technology and skipped the necessary step of mastering chat with a human agent first. When customers engage with a company through chat, they don’t want to come across an algorithm trying to guess what it is that they want. And yet it was happening all the time, turning chatbots into a source of frustration.
Nonetheless, this doesn’t mean chatbots are doomed to failure — on the contrary. Chatbots will provide customers a channel where they feel they don’t expose themselves too much, which might happen over the phone, but it’s interactive enough to make them feel like they can say what they have to say and get an answer right away. It’s the perfect balance between the immediacy of voice and the relative anonymity of email.
The important question now is: How can you make chatbots in customer service work?
First, you get your team to use chat to reply to your customers and analyze the interactions. What kind of questions are people asking? How are your agents replying? What can be automated? Then, you program your chatbot according to that analysis and your business needs. Train them to answer your most common question by connecting them to your FAQs and to direct customers to an agent for the more complex issues.
Again, similar to the cases I mentioned above, it’s all about the perfect balance between humans and machines.
Multilingual customer service
There’s been a lot of talk about machine translation (MT) in customer service. However, most customer service leaders are still skeptical about introducing this technology into their workflows.
Over the past few years, I’ve been leveraging MT in a number of different contexts, from professional translation agencies to huge e-commerce companies. But it was only when I joined Unbabel as Director of Applied AI, that I realized machine translation will play a major role in the future of customer service.
MT will be an important component of many enterprises’ automation strategy. It will make replying to tickets more efficient, improve the customer experience, and help customer support agents write in multiple languages even if they’re not native speakers.
But will Machine Translation for customer service work?
Well, we all know that we can’t yet trust MT to accurately translate all customer interactions. Even the most advanced Neural Machine Translation (NMT) systems aren’t necessarily up to the job of delivering the consistent tone of voice and near-total accuracy that modern enterprises require, especially when it comes to branded language and customer terminology.
So what these machine translation systems are missing is human oversight. You just need to strike the perfect balance between human expertise and artificial intelligence. NMT combined with advanced, automated quality assurance and post-editing by humans, ensures translations that are sound and native-sounding — and often delivered in a matter of minutes.
Sometimes, in our quest for higher customer satisfaction, we overlook the people who are ultimately driving scores higher: human agents. By giving agents the best workflow imaginable, along with the time to focus on the complex queries that require their unautomatable empathy, we can create a better workplace and better support outcomes. To make that happen — to give agents their most creative workday and customers their answers as quickly as possible — we’ll need to turn to Artificial Intelligence.