The following article is a guest post written by Edmund Ovington, VP Global Alliances at Unbabel, a company that combines Artificial Intelligence and human expertise to remove the language barrier between companies and their customers. For more information about Unbabel visit their website.
When people start talking about automation in customer support, and how computers are taking over and why we no longer need human agents, I think of my dad.
This may not come as a surprise to you but my dad hates automated things and computers (Little Britain is likely to blame). He was, of course, born in a world where you would speak to bank managers, toll collectors, and supermarket cashiers – in fact you’d often have a personal relationship with them.
And he’s not the only one.
They say humans yearn for human contact and, if given the choice, they will always prefer interacting with another human. The logic behind this is that we are a highly social species, we live in society and are innately designed or programmed to interact with each other.
This argument has some scientific truth to it: in one study by Ryan W. Buell, customers who used ATMs more than human tellers had a lower level of satisfaction with their banks. Buell, who is now at Harvard Business School, found in a different experience that people calling Metlife to claim death-related insurances, receive an automated condolences message – and were a bit freaked out.
On the other hand, wasn’t this argument of human contact used against removing toll operators? Or even the iconic phone operator? Against self-service checkout at supermarkets?
The thing is, not all customers are like my dad. Personally the less humans I need to ‘deal with’ in everyday ‘normal’ tasks, the better. If I can solve a problem or get something done without talking to a human I feel pretty productive and thankful about it, actually.
In other words, this is what we all need to understand about automation in customer support. You’re automating it for the people. And it can’t be feast or famine. It needs to be just right. It needs to be nuanced.
Since I joined Unbabel as the VP of Global Alliances, I’ve been working closely with companies like Zendesk, Salesforce and Concentrix, to figure out ways to improve customer experience and make it more efficient.
And when it comes to automation there’s one rule of thumb: Automation is great, if – and only if – it brings you closer to your customer. Anything else is a bad move.
So how do you know what to automate and what not to?
1. Understand your product
It all comes down to your business and product. Is your product easy to implement? Or do you usually have to walk your customers through implementation? Are you B2B or B2C? How’s your customer service team organized? Do you have physical stores or is everything online?
These are all relevant questions you should be asking yourself at this point. If you have a complex product you need to make sure that automation doesn’t stand in your way. You want it to be the source of simplicity and efficiency not of frustration and negativity.
A good example that I like to give is Bang & Olufsen, the Danish high-end consumer electronics company that designs and manufactures audio products, television sets, telephones, and so on. So basically their traditional customer base is people walking into the store and talking to the person they bought the product from. I mean, if you’re going to spend 10k-20k in a TV set you might as well talk to a human to make sure you’re making the best choice and that you have excellent customer service in the future.
However, more recently, Bang & Olufsen diversified and released a new line of products called B&O Play where they are selling headphones and speakers to a younger audience at a more affordable price. So, in this case, they had to flip their model. These new products are simple, easy to use, and highly scalable, and their target audience loves automation. In the end, to deal with scalability what they did was to keep providing excellent customer service by having human agents while using AI to make those agents much more efficient.
So, in the end, it’s all about the product you’re selling and who you’re selling to, which brings me to the next point.
2. Know your audience — age and geography
It’s crucial that before you look into automation you get to know your customer. How old are they? Where are they from? What kind of help do they need while using your product?
Most people agree on this: if millennials are your audience, automation is great. They’re not that much into talking to other humans. And they hate talking to anybody on the phone.
But does this mean they want everything to be automated? Well, not necessarily. And here’s where most people get it wrong. It’s not that Millennials don’t want to talk to you, they just want a better customer experience overall, and if that means less of hassle they’re gonna go for it.
Another thing you should bear in mind is looking at where your customers are coming from. Most people neglect the importance of culture in consumer behavior, and yet it is paramount.
Do people in Japan like chatbots? Or do they prefer talking to a human? What about in France? Or in the US? According to a survey by LivePerson, for instance, European countries tend to be more receptive to chatbots. But people in the US? Not so much. 59% of Americans said they preferred talking to a human other than a machine.
Bottom line is: get to know your customers, and how old they are and where they are from to deliver the best, tailored customer experience.
3. Analyze your common queries
The truth is, your customers are used to a certain standard and they know what to expect. So make sure you look back and understand the kind of help they need from you. What type of questions do they ask? Are those easy to solve problems? Things you can automate or put in your FAQs?
Look into your common queries and if you have a lot of those there’s just so much you can do to make the whole process better. There’s a lot you can automate and improve. So why not put part of your customer service operations on autopilot?
That’s exactly what platforms such DigitalGenius are doing by using AI to understand conversations, automate repetitive processes and solve your customers’ problems. This will help you automatically deal with the most common queries such as refund requests, order status inquiries, cancellations, and so. And the great thing about this is that it will allow you to cut down costs and keep things efficient.
4. Automate what you can
In the end, it’s not a matter of artificial intelligence and automation replacing your customer support agents but rather giving them superpowers for them to be more efficient.
It all comes down to making smart decisions and understanding where automation can help you achieve better results, increase customer satisfaction, cut down costs, scale operations, and have a better and personalized customer experience. Live chatbots sound great. Self service options are super useful. AI-powered translation too. I mean, your possibilities are endless.
It’s really just a matter of knowing what will bring you closer to your customers. And who knows, even my dad may change his mind.
Edmund Ovington
VP of Global Alliances
Edmund drives value for Unbabel’s customer base through close relationships with key partners and investors (including Salesforce, Microsoft, Zendesk and Concentrix), as well as building out Unbabel’s US operations from NYC. He has over 10 years experience scaling technology, specifically B2B SaaS, companies such as Yammer.
