Contact centers handle customer service issues by routing incoming phone calls, emails, text messages, and software tickets to relevant experts who can answer questions and help the customer with his or her problem. The “expert” portion of this process still requires the human touch, but much of the routing and categorization of tickets can and should be handled through automation.
Interestingly, contact center automation isn’t limited to routing and categorizing calls. As you might imagine, the portion of the phone call or email process that happens after a representative answers is the most crucial part, and can also benefit greatly from the assistance of intelligent automation. A human employee would have to manually search for the client’s account and read through the history, whereas AI can do that automatically and highlight the most relevant information. Done right, the automation of customer service contact requests cuts down on the mental load experienced by call center representatives, helps customers find answers more quickly, and generally leads to a better and more efficient experience for both parties.
We’ll get into the specifics of how automation improves each process later, but first, let’s clear up a common misconception about contact center automation, and clarify some proven benefits.
Before we discuss the benefits in detail, first let’s address the elephant in the room.
Some attempts at call center automation use robotic answering machines to field customer calls and connect to human customer service specialists as needed. This is a basic implementation automation, where the computer handles repetitive or tedious tasks so humans don’t have to. English speakers can press “1” for example, while callers who prefer Spanish can press “2.”
Robotic answering systems save the company time because employees only receive calls that have made it past the robotic screening process; however, they don’t do much to make customers feel welcome.
In fact, these robotic answering machines have become somewhat notorious for turning already frustrated callers into outright angry callers who just want help with their problem. Getting lost in endless menus and repeatedly pressing “0” to speak to the support team is no one’s idea of great customer service.
Thankfully, advanced call center solutions are so far distanced from these rocky and primitive attempts that customers should have nothing to fear by dialing your support line and interacting with an intelligently automated system.
The application of artificial intelligence (AI) and the use of business process management (BPM) are what make the difference between a clearly computer-driven interaction and a helpful, streamlined customer service experience.
To illustrate how AI- and BPM-assisted automation not only eliminates these typical call center issues but can actually make the experience a pleasant one, let’s look at some examples.
When a customer calls or sends a message, they usually have a specific goal in mind. They might want to order a new part, get tech support on a confusing product feature, or speak to a manager, for example. Each of these goals requires a different department or employee to answer the phone. A customer could get bounced around to several representatives, often waiting on hold for some time in between connections.
The last thing a frustrated customer wants to do is wait for a long time to have their questions answered. Waiting on hold only to be routed to the wrong department a few times is likely to leave a bad impression of the company as a whole and might even discourage the customer from doing future business with that company.
Automation reduces average handle time by ensuring calls, emails, texts, and other customer contacts get routed to the right person immediately. Additionally, chat bots can often answer simple questions or help customers find more information on a particular topic. Competent bots can solve the problem instantly and free up human workers to solve more complex problems that require critical thinking.
If a customer does want to speak with a representative instead of asking a chat bot, the AI running in the background should be able to pull up that customer’s account information so the caller doesn’t have to waste time explaining who they are, what product model they have, and other information the company may already know.
When customers send support requests in email or text message form, they often include plenty of details and maybe a few examples of the issue they’re experiencing. While that’s helpful to human readers, it can confuse bots without human-like intelligence that can’t piece together enough context to understand the main problem or how the customer feels about the situation. This happens during phone calls as well, but it’s particularly prevalent when a person tries to type out enough background information to describe a problem.
The person may say something like, “The plug to my coffee maker got bent out of shape, and now it doesn’t fit into the outlet in my kitchen.” A non-AI bot may pick out a few key words such as “coffee maker,” kitchen,” and “doesn’t fit” and mistakenly provide a link to an article about creative ways to store your coffee maker between uses so it doesn’t take up space on the kitchen counter.
As with many other scenarios, this is a case where RPA bots are limited, but AI would be able to discern the context of the customer’s email and understand what the person is actually trying to say. Instead of offering interior decorating suggestions, an intelligently automated system would recognize the real issue and direct the customer to someone who can help them order a replacement cord with a new, unbent plug.
Experiencing a problem with a product is bad enough; it’s even worse to be forced to spend time seeking a solution. But not all customer support requests are about product issues or similarly frustrating problems. Sometimes, all the caller wants to do is check on the status of an order or confirm that a returned item was received on time. While these relatively happy callers take up time in the queue, those with an urgent issue get more upset with every passing second.
On the phone, tone of voice and manner of speaking reveal everything about how a person might be feeling. Humans can tell when another person is calm and happy, slightly irritated, or outright angry. In fact, you can probably make an educated guess about the emotional state of the other person even if you are simply emailing or texting with them. Short, clipped responses simply feel different than longer, more detailed messages. Plus, it’s a dead giveaway how someone must be feeling when they use emojis or type in all caps and use profanity.
These things seem obvious to us because we’re people, but simple call center bots usually don’t have the programming to pick up on emotional cues. Therefore, an irate caller gets routed to the exact same place as a kindly old lady wanting to ask about store hours on Sundays.
With the addition of AI, automated systems can run a sentiment analysis to categorize customers based on their emotional state. This can prepare the customer service representative for potentially difficult interactions. In some cases, it may also be helpful to route angry customers to managers or other high-level employees, automatically escalating the support request to bypass call center employees entirely.
In a typical call center setup, employees answer phone calls or open email messages without an intelligent system assisting them. They often go in blind, with no warning as to the customer’s mental state and no idea of the problem. Therefore, the employees must come up with solutions on the fly. No matter how competent a call center representative is, they’re only human. It’s only a matter of time until they get flustered by a difficult customer, misunderstand a question, offer a solution that’s already been tried, or even accidentally disconnect the call and have no idea how to reach out again.
By the time an employee picks up the phone, an AI automated system should already have processed a number of things. The computer has identified the problem and the emotional state of the caller. The system has recorded the caller’s phone number, pulled up the associated account information based on that phone number, and provided a full history of purchases and previous customer service interactions to the representative. If the system notes, for example, that this caller has already contacted the company for help three times and seems irritated, it can warn the representative of these considerations and make recommendations based on those facts.
An intelligent system can also provide suggestions for solutions that don’t appear in the customer’s history and therefore may not have been tried before. Essentially, the AI can act as a guide to help the customer service representative make efficient and logical decisions based on all observable data.
By the time a customer calls or emails about a product issue, they may have already sunk a significant amount of time into trying to solve the problem on their own. Several other factors also tend to increase frustration levels:
Each of these factors can be eased by the intelligent automation solutions we’ve discussed. Instead of getting hung up on a particular part of the process, the customer could experience a relatively quick and seamless connection with an expert human who, guided by AI, picks up on the customer’s feelings and uses the on-screen suggestions to offer the best solution possible.
Less frustration makes for a more pleasant interaction for both the customer and the representative, and puts the company into a much better light—even if a problem prompted the call in the first place.
We can’t speak for every company that has tried to implement some form of automation in their customer service process. After all, some of those efforts at automation have led to the infamous robotic answering services we discussed above.
What we can do is highlight what our clients have achieved by using AI- and BPM-assisted automation systems built by Nividous.
One company, for example, previously needed employees to monitor support tickets 24/7. This team also had to respond to each message and update all support tickets by hand. Nividous was able to automate their entire support desk by using bots to monitor messages and send automatic and timely responses. Now, each customer sees how their problem is being addressed by the expert team working behind the scenes without having to call or email a customer support specialist at all.
As another example, an insurance firm that handles over 1,500 support requests every day needed a better way to manage the complex tasks of categorizing and solving such a huge volume of tickets. AI-assisted bots designed by Nividous now handle each request, analyzing, classifying, and performing sentiment analyses on every email. The bots also update each customer profile with the most current information and intelligently respond to each email using a template created for the relevant customer category. With bots taking care of these details, human employees can spend their time completing other useful tasks.
Numerous other client success stories can be found here.