You pick up the phone. Palms sweaty, you punch in the numbers.
“Thank you for reaching Comcast. Your wait time will be approximately 13 minutes. Please hold.”
Your suddenly bright day turns into Dante’s Inferno.
No amount of preparation could’ve prepared you for the storm of frustration brewing within you. Exasperated, you slouch back in the couch. An existential crisis is looming. After the rapture and the resurrection of Walt Disney occur, a glimmer of hope…
“Thank you for your patience. The call center is experiencing an abnormally large call volume right now. Your new wait time is 27 minutes.”
27 minutes. The time it took Yigrem Demelash to run the fastest 10k in the world. The time it takes to employ the zen approach of counting to 100, 16 times.
The reason call centers and customer service are such a huge cost.
The make or break for a good user experience.
The key determinant of how customers view a company.
When users want information, they want it fast. Studies have shown that 62% of B2B and 42% of B2C customers purchased more after a good customer service experience. 69% attributed their good customer service experience to quick resolution of their problem.
What if there were an omniscient digital entity that was able to respond to user requests at lightning speed, all at once?
Specifically tailored to the needs of any business? Check.
1,000 times cheaper than a call center? Check.
So why now?
Machine learning and natural language processing have come a long way in recent history. The time for parallelizing and automating complex tasks is now. Almost any current computer or program can harness state-of-the-art algorithms much more advanced than the ones that powered Deep Blue – the machine that beat the best chess player in the world 20 years ago.
Promising use cases
There are multiple industries where Chatbots will be useful. The most common use case for chatbots is customer service. They can often eliminate the need for human involvement since most of the time customers are going to be asking the same questions over and over again. These questions can easily be answered by a chatbot.
Imagine you want to go on a vacation but don’t want to do that much research. Chatbots are really good at making decisions based on previous decisions, which are also known as ‘decision trees’. In this case, a chatbot could know exactly what preferences you had from previous decisions and/or previously answered questions and make intelligent decisions based on that previous context. This gets you from point A to point B much faster. Integrate that with a chatbot that knows the weather and you can easily plan amazing trips for the immediate future.
Want to find out information about your favorite sports team? Or even the standings for the whole division? Chatbots allow you to seamlessly switch from one piece of information to another, instantaneously and within the same context. This way all the user has to do is scroll back up instead of making another Google search.
Nothing is quite so therapeutic as a good listener. What robots can achieve perfectly that humans cannot is ‘Unconditional Positive Regard’, which is basically treating people consistently without judgment. This humanistic approach to helping allows people to confide in something and overcome emotional hardship without having to pay for expensive social services.
Here are some companies that need chatbots:
With the recent proliferation of men’s fashion, many novices are looking to increase their sense of style. However, there is no single hub of information or person to ask for tailored recommendations. GQ could pioneer the chatbot movement in fashion with their strong digital presence.
This one’s anecdotal, but I recently had to go through 4 different people and wait for a total of 40 minutes just to inquire about a transfer. Obtaining this information should be much easier to do and require fewer workers.
The premier platform for buying PC games. Unfortunately, their customer service is basically nonexistent. Having a chatbot to settle purchase problems will help retain customers.
I’m Afraid I can’t do that Dave…
It’s important to identify the possible drawbacks of Chatbots and to implement functionality that will mitigate these areas.
- Bot Personality is very important. If you don’t give a bot empathy and proper grammar, it’ll quickly turn into robotic Al Roker. You don’t want robotic Al Roker.
- Chatbots aren’t well-versed in expletives naturally. Incorporate some ducking good curse word filtering.
- Language support. Completely understanding what a user is saying in every language is difficult because different languages - even dialects - have different sentence structure.
Tips for Success
While in many ways chatbots are still in their infancy, there are some key best practices that product developers should follow:
- Figure out your top users’ needs. Target them by making a chatbot that solves these needs in a robust manner. Then, focus on the edge cases and other pieces of information.
- Create a whole conversation. Have some hellos and goodbyes. It needs to feel human and personal.
- If you wind up using an icon to represent the chatbot, make sure it’s not too creepy or human-like. Always remember, the uncanny valley exists.
- If you have a chatbot, you’re understandably pretty excited. However, don’t let that carry over into a decision that negatively impacts user experience, like popping up a maximized chat window that is ¼ the size of the screen on the web.
- It’s important to establish a good product roadmap and thoughtfully assign resources to build the chatbot, whether that be in a natural language processing service, design, development, marketing, etc.
The Main Message
Heh. You get it? It’s like a joke about chatbots sending messages, and I’m sending you a message right now. In the digital space, just like a chatbot would do. A ‘play on words’, as some would say. Anyway, chatbots can revolutionize the customer experience for a wide variety of organizations. Leveraging the use of modern natural language processing techniques, chatbots can solve problems like alleviating call center pressure, helping those in emotional need, and creating deeper connections with customers.
Kevin O'Leary is a Developer / Product Strategist at Yeti. Sometimes knows stuff about stuff. Mostly doesn't.