New technologies, such as chatbots and artificial intelligence (AI)-enabled communication make around-the-clock customer service a reality. It’s just another way digital commerce is evolving to give the customer what they want. That’s because 63% of people are okay with using chatbots to interact with companies and brands and 33% prefer to use chatbots instead of speaking with a person when making a purchasing decision. No wonder MarketsandMarkets estimates the chatbot market to be worth $3 billion by 2021. But for companies that place a premium on excellence in customer service, are chatbots killer technology or the kiss of death? Like all technology, chatbots have their place. Let’s see where that may be in your organization.
Exactly What Can a Chatbot Do?
Chatbots are software programs that can hold a conversation with a human. They appear smart, but they have their limitations. They are not exactly the AI we are used to seeing in movies yet.
Chatbots are built on either artificial intelligence or decision-tree logic. With either approach, chatbots still depend on humans; either to teach them or program them properly. Chatbots built on AI accept freeform input and they learn as they work provided humans correct them when they make an error. Chatbots built on decision-tree logic look for key input words to guide their response. When deployed properly, chatbots can enhance customer experience. When pushed beyond their limitations, the results can be disastrous.
Effective Chatbot Applications
Chatbots can be great applications for simple customer queries. A human customer service agent can only help one customer at a time. Chatbots are masters of multi-tasking and can help many customers at once. Chatbots don’t need to sleep, eat, or take a bathroom or smoke breaks. The don’t form unions, file grievances, or sexually harass their coworkers and they are never late for work or get sick. Sounds like the perfect employee, right? They can be if given the right job. Chatbots are great for:
- Helping customers make eCommerce purchases
- Answering customer questions about products or orders
- Assisting live customer service agents with support
Chatbots can save companies money on staff and benefits. They can also frustrate customers if they aren’t deployed properly.
It’s important that chatbots identify themselves to users up-front. This sets up realistic expectations for the human involved in the conversation. Because, like children, we tend to be more patient with computers. 75% of people want to know they are interacting with a computer, so you should tell them 100% of the time. Balance your desire to have chatbots appear as human as possible with the need for the transparency that is vital to retaining customer trust. Also, don’t forget that chatbots operate as part of a larger ecosystem and are just one touchpoint between the customer and your brand. To be effective, they cannot be restrained to one silo and must integrate with your CRM and other systems.
In Designing Simple Chatbots for Customer Service, Tom Griffiths recommends that you only embark on chatbot implementation after you have collected enough data to allow you to intimately know your most common customer pain points. These are the places chatbots can help and free your human customer service resources to focus on the situations where human intervention and a personal touch is required. For example, an FAQ chatbot can answer most simple product and order inquiries while live customer service representatives solve complex problems and sales representatives focus on providing personalized solutions.
Chatbots for Customer Service Examples
When Allstate made a major change in its small business line, agents and customers had questions. So many questions that their call center couldn’t keep up with agent and customer inquiries. Allstate deployed an AI virtual assistant they named Allstate Business Insurance expert (or ABIe). This chatbot easily handles questions from 12,000 agents and customers and paid for itself in one year.
Real estate agents are using Apartment Ocean to increase revenue and reduce expenses by fielding customer inquiries. When a lead lands on an agent’s page, the chat starts using Facebook Messenger. The chatbot collects all of the information the agent needs to determine the viability of a lead and frees the agent and customer service to focus on closing deals instead of answering common inquiries.
Starbucks has been a leader in technology when it comes to enhancing customer experience. They were pioneers in providing free wi-fi and digital reward cards. No surprise they’ve incorporated chatbots in their app to make ordering as easy as possible. Just speak or text your order to the chatbot and it will tell you when to pick up your order and the total cost.
As you can see, the common thread in successful chatbot applications is that they handle routine, predictable tasks and inquiries exceedingly well. Chatbots can quickly provide answers to simple questions like order status, pricing, and routine tasks like how to update billing information on an account. But it’s important to maintain a personal touch. Humans must monitor chat sessions and take over when necessary. Chatbot fails can be the kiss of death. Remember, every customer experience has elements of value creation (positive value) and value destruction (negative value). The value destruction multiplier is not one for one. That means it takes far more positive value to offset any negative value. If customers feel they are getting the cold shoulder from a chatbot, it reflects poorly on your brand and can reduce customer retention.
Are Chatbots in Your Future?
Despite all the hype about chatbots and AI, human customer service and sales representatives still have jobs for the foreseeable future. It will be years before AI can handle complex problems and human interactions that require empathy. But that doesn’t mean that chatbots don’t have space in today’s customer service. When it comes to handling the routine product, account, and order inquiries a chatbot just might be the right (artificial) man for the job.