Customer Service Automation’s Dirty Little Secret
Everyone knows it’s “five times easier” to keep a customer than it is to get a new one. Everybody knows customer loyalty is important to revenue. Everybody knows that a positive customer experience can help generate growth. Do decision-makers really act in accordance with these beliefs? Customer service automation trends make it clear they don’t.
In order to see this in action, all you have to do is contact customer service of the world’s biggest, most well known and well-resourced brands. Every single one of them relies on a high degree of automation. That’s because customer service automation is always sold as offering the ability to provide instant access to answers and fast problem resolution. In reality, all those solutions really accomplish is “deflection and containment”.
Every time someone closes a live chat or hangs up a phone, that’s considered a “contained issue”. Some vendors even refer to “digital containment” as an optimal strategy. It’s like saying: “Assume your customer is on a digital channel because they’ll never want to talk to a person and… let the rest take care of itself?”
If you think about the times you’ve been a “contained issue” – meaning when you’ve hung up the phone or closed a chat with customer service – how many of those times did you find yourself truly satisfied? Many of those times, you were probably annoyed, frustrated, irritated or even outraged, tired and simply had something better to do with your time. The problem for brands, again, is that in many cases they treat that outcome as success.
Customer service automation often treats customers like a nuisance, focusing on saving time and cutting costs rather than offering customers the best possible experience. Those experiences most often require human empathy, immediacy and understanding – adding machines to this process, unless it’s a really low-intensity issue, just makes it seem like the brand doesn’t care.
Ultimately, customer service automation can’t replace the effectiveness of human agents for most brands, nor the need for customers to feel cared for. Fortunately, there’s hope: you can focus on the best human-centric practices in customer service, while the technologies that are most often used for this type of automation can actually give you even better customer engagement outcomes when applied to marketing and commerce.
The More Things Change, the More They Still Piss You Off
Many consumers began to despise customer service automation upon their first encounter with automated phone systems (Interactive Voice Response, or IVR for short). Many customer service leaders thought live chat was going to fix everything. It was believed that an “asynchronous” chat could help you activate a credit card, submit a complaint or respond to a notification with less cost and frustration than going over the phone (and, again, dealing with an IVR).
The thing is, this still doesn’t help you do things like: rebooking a cancelled flight on short notice, filing an insurance claim when your car’s been totalled, putting the lights back on when the power’s out or helping you find a gift that didn’t get delivered. It doesn’t get a product working when you need it working now – human agents, their voices and proactivity are the best weapon brands have to resolve these issues in a positive way.
Everyone’s gone through an urgent customer service issue and seen automation fall flat at one time or another. While phone lines happen to get disconnected, live chats can’t escape the “are you still there?” Odyssey either.
So, aside from a couple situations, chat hasn’t solved these kinds of problems and, for the most part, a chatbot won’t help much more. Many brands, for instance, think an AI assistant can help them realize all the potential they see in customer service automation. One of the biggest benefits is sold as instant answers, but with critical issues, there isn’t enough margin for error to allow chatbots to significantly improve the experience. Brands focused on this outcome are being distracted by the dollars they think they can save. What they’re not seeing, of course, is the dollars they’ll never earn from the customers who ultimately leave them.
Improving customer experience has been shown to drive revenue growth. This doesn’t mean automating it – this means making it the best experience it can be and using customer service automation only where it actually helps contribute to that goal. This includes new AI solutions, no matter how attractive their promises may be. Otherwise, you’re just going to end up following the same flawed “deflection and containment” strategy that brands have in the past, only with its downsides “supercharged” by the efficiency of AI.
Avoiding “Deflection and Containment” on AI
Tim O’Reilly, a longtime leader of the web and (full disclosure) an investor in Automat, often says that the approach of automating “people first” is too simplistic. It’s the most basic way you can think of how to use AI. What AI really offers is new functions, roles and opportunities to enhance the experiences we’re already used to, or make it 10 times better with a little help. You can’t do that with customer service automation, but you can do it by adding interactive AI to customer acquisition.
It’s expected that by 2020, 85% of customer service interactions will be managed without a human agent, and that AI solutions can automate as much as 75% of an individual customer service agent’s time. Time-saving and, by extension, cost-cutting can be a compelling benefit of automation. But, viewing customer service as a cost-center also avoids the fact that it’s actually an opportunity to actively retain and upsell customers. The prerequisite here is that this can only happen once a customer’s urgent issue has been adequately addressed.
For example, “ticket deflection” might work for a lot of apps or a digital services. On the other hand, with physical products or more expensive services, designing systems to avoid helping consumers directly can lead to negative results. Automation can improve efficiency, increase self-service usage and make it easier to meet SLAs, but these kinds of experiences can make a customer feel as if they’re being treated poorly and magnify any underlying frustration that causes them to seek out service in the first place.
While it could be argued that the faster your service is, the more likely you are to meet your metrics, this isn’t an argument for automating customer service – this is an argument for changing the metric. Cost-cutting for it’s own sake usually just sacrifices opportunities to grow. Just because customer service can seem repetitive doesn’t mean you won’t benefit from making your customer service a better experience, rather than a cheaper one.
Realistically, you’re only going to get 5-15% automation in any customer service scenario. The only things that should be automated are the things where customers don’t want to actually deal with a person – getting a refund, checking the status of an order, triage before talking to a person or any sort of one-turn interaction. If someone is gravely injured overseas, has a flooded basement, needs an essential service up and running – these are all situations where you need real care and acknowledgement from another human being.
Customer Service Needs a Human Touch, But Automation Can Still Help Elsewhere
T-Mobile is one example of a brand that has actually abandoned the traditional customer service automation approach for these very reasons. They recognize that customer service automation can have limited benefits, while customer service is often a delicate task that needs a human touch while today’s consumers would rather get care from a human being – not a machine.
In terms of providing care, humans have a clear role. People want immediate responses from representatives who are skilled, knowledgeable and empowered. This is something you expect when you have an urgent, nuanced problem, but something you almost never get when you do something like, say, shop online.
Brands have acted this way for a long time. Online shopping has always been “self service”, but what if there was a way to make it more than that? When we shop we face complex questions, both about our needs and the product details that address them. Guided shopping is a perfect example of a place where automation and AI can enhance the experience and exceed a customer’s expectations – not make a bad experience worse.
The “75% time saving” for individual agents advertised by companies like IBM can be attractive, but automating entire customer service interactions increases frustration too often to be truly useful. Consumers looking for answers can’t always get the exact ones they need. If there’s urgency or any sense of irritation underlying their issues a human hand will always be needed to talk them through it or simply to hear them out.
Customer Service Automation Can Replace Agents, But Only So Many
If you can effectively automate 5 percent of your customer service and you happen to run one of the few contact centers that has more than 5000 agents, then you can clearly benefit from customer service automation. That’s a good 250 agents you can reassign. If you have a contact center of less than 20 people, however, how much cost-savings are you really going to create and still have enough people engaged in feeling your customer’s pain?
Customers do have frustrations, but most of them – where not limited by the absence of a live person – come down to service interactions not helping to resolve their issues, representatives lacking the knowledge to help them, or having to repeat information multiple times when providing information to an agent.
In these cases, the right systems can all serve to better inform, empower and enable the customer service agent to do their job, and allow the customer to access an agent faster and more quickly than otherwise possible. This is the kind of automation that gets things right, but it doesn’t automate “75%” of customer service interactions. It actually puts human agents at the center of the job, amplifying their ability to communicate directly with a customer in ways that offer the best possible experience.
If your brand doesn’t have a massive contact center, the thing you need to think about isn’t customer service automation – it’s finding more ways to add automated service (and a little bit of care) to your revenue generation efforts. This is where you’ll see the biggest benefits.
The Real Way to Automate Care People Still Think Machines Can’t Care
A deeper understanding of robot emotions would lead one to understand that machines can, in fact, care. However, consumers aren’t there yet – 34% of consumers report that not being able to reach a live person for support is the most frustrating aspect of the customer service they receive. Situations where knowledge and intelligence is needed without urgency and care can actually yield better results from automated assistants and advisors. If you add machine assistance to the product discovery, consultation and shopping experiences they get, you can also increase the overall satisfaction consumers have.
This kind of guided shopping helps people feel like VIPs, and doesn’t require you to program your chatbot to feel pain. Customers want reliable experiences that meet their expectations, enjoy experiences that exceed their expectations and resent those that fall short. New automation capabilities don’t let you replace your customer service capabilities, but they can still generate more revenue and make your customers happier. Isn’t that what you want in the first place?
Where to Look for Automation Outside of Customer Service?
Brands need to consider the right channels where automation can engage consumers directly and in a way they’ll love. Conversational AI for marketing and commerce, along with some support functions, can revitalize a brand’s relationship with consumers by offering one-to-one personalization throughout the entire customer experience.
It’s also something that, over time, will make it easier for consumers to understand that robots really can care. In the meantime, brands can see benefits like multiple minutes of engagement per interaction, higher customer satisfaction, and 30% or more sales lift when compared to traditional ecommerce experiences.
You don’t need to water down your customer service experience – think about the opportunity you have to improve the shopping experience you provide instead. In maximizing profitability, the answer is clear: don’t lower the bar on customer service, raise the bar on customer experience from the first moment a customer engages with your brand.