What Is Required To Make Chatbots Succeed

A few months back, I wrote a blog on how much the “Live Engagement” space, now known as “Conversational Commerce” has evolved over the last six years. To keep the conversation going, here is the next installment on chatbot success. In the interest of keep these reads short, I will be breaking this into two articles:

  • This post highlights the current business circumstances and the consequences of what we are seeing with chatbots today
  • My next article will feature how Sitecore and their ecosystem of partners can take chatbots to where they need to be.

Let's start with the basic definition of a chatbot as context:

chat·bot a computer program designed to simulate conversation with human users, especially over the Internet.

The start of the chatbots evolution was a bit rocky early on, with headlines from a couple years ago reading “Over-hyped”, “Frustrating experience”, “Not easy to implement”, but over the last year this is rapidly starting to change.  Here is my take on the strategic intent of chatbots, the top contributing factors to the current situation and the corresponding implications to address to make thing better.

Chatbot Strategic Intent:

In order to simulate conversation with human users, chatbots need a customer experience and, ultimately, a brand experience that provides businesses the right solutions and information to create personalized captivating experiences for these customers.

Top five supporting determining factors and corresponding implications for chatbots:

“I thought I was having a conversation, but it felt like a bad Interactive Voice Response experience that I get on the phone all the time”

If a customer asks a question and the chatbot gives them a list of predetermined answers to choose from, that is not a conversation. There is a need for personalization to make it conversational. Businesses need their chatbots to understand customer history, personalize and contextualize beyond basic FAQ responses. In order to accomplish their personalization goals, businesses will need to leverage artificial intelligence (AI) and automation chat technologies, relying on chatbots, machine learning, predictive chat and routing capabilities to succeed.

“The chatbot on your homepage is intrusive.”

I just experienced this last night on a car dealer website resulting in a bad experience for me. “How can I help you today?” they interrupted repeatedly. I was not looking for help, I was just doing some research! A better approach would be to “micro-target” where you utilize the chatbot based on customer behavior and intent. For example, on retail ecommerce sites chatbot conversations are being initiated based on the time spent on a certain product detail page for complex or fashionable items to drive the customer to conversion; Items that require support to make the sale.  This is allowing chatbots to not just provide operational benefits like decreased costs and more productive agents but quantifiable impact to top-line revenue. These sessions often start with a chatbot and transition to a live click-to-chat session. Chatbots won’t work in all customer circumstances; there are situations that require a true person’s touch to solve. Don’t fall into the trap of thinking that your chatbot will eventually be able to answer every query your customers have. You can design chatbots in a way that they pass requests to a human agent including the context they might need to continue the conversation. For these situations, you must ensure that the handoff from chatbot to person is as seamless as possible. 

“It could not answer my question and told me to email or call support!” 

When a customer attempts to interact with your brand through a specific touchpoint or channel, there is a reason. It may be their preferred communication channel; driving a customer to another touchpoint they didn’t choose is less convenient and extremely frustrating. In today’s always on world, the customer expects consistency, regardless of touchpoint, and the chatbot solution must provide this.  Measure results from your chatbot and augment with customer data from other business systems, such as your CRM to help deliver persuasive interactions with your customers. Powerful AI solutions can leverage customer background information to modify each interaction to that customer’s specific situation.

“It just did not get what I was asking.”

If you are going to take the time to make your chatbots available to your customers, you need to make sure it can understand and react to what your customer wants to accomplish. Use a solution which understands the natural ways in which humans converse dynamically versus deploying a prescriptive chatbot. Accurate Natural Language Processing (NLP) is central to bringing the best experience. Set up your chatbot to never stop learning and improving. Just like a child, chatbots need the ability to continuously learn the tones of language and adapt based on what they’ve learned.

“I ask a simple question and the reply is ‘Please select from an option we can help you with.’ How does that help me?”

Without a clearly articulated purpose, your chatbot could do more harm than good when it comes to customer approval. It has the potential to be an accommodating tool, provided that your customers have realistic expectations, are aware of what the limitations are, and when they’ll be carried over properly to a live agent. With a chat interface, there is an expectation that a customer can have a conversation with either a chatbot or a human. Removing both of those is a recipe for an angry customer. Even if your customers know they’re engaging with a chatbot, they still want a human-like experience. Look at your current customer service representatives’ scripts and interactions to define what conversation flows should look like. Taking the conversation flow into account will provide the best experience for customers. 

If you implement a chatbot without a business objective in mind, there’s a real chance you’ll just end up wasting time and investment on a technology that doesn’t deliver and even worse creates such a bad experience that your customers might leave. Businesses with clear objectives, that invest in the right chatbot platforms, will be able to offer a high degree of personalization, leading to satisfied customers and better brand experiences. To enable these improved experiences, chatbot platforms are critical in making this personalization happen by leveraging customer, product, and interaction data that lives across the organization. Chatbot platforms must be easily integrated and offer business-friendly tooling to allow CX professionals to craft the experiences their customers demand. This will be the topic covered in my next post as I look at how Sitecore and their partners tackle these issues and address these implications. Stay tuned.