Why chatbots in the contact center won’t replace live agents


Share on LinkedIn

There are more than 34,000 chatbots on Facebook messenger alone built by brands to help people access anything from news stories to suggested cheese pairings for wine. Gartner predicts that by 2021, more than 50% of enterprises will spend more per annum on bots and chatbot creation than traditional mobile app development. Long story short, chatbots are en vogue and they are everywhere.

The omnipresence of chatbots is a result of two primary factors, advances in AI and the emergence of messaging as a preferred means of communication.

For customers who expect faster solutions around the clock at scale, chatbots are a great alternative to traditional IVR systems. The time commitment needed to get through the complex maze of menu options in an IVR system, coupled with short attention spans and the culture of instant gratification, make chatbots a great fit for a variety of use cases. Smart Insights has labeled real-time interaction as one of the four major trends that will shape the retail industry of the future.

Chatbots in the contact center

We know that contact centers play a pivotal role in overall customer experience. So, should contact centers invest thousands of dollars in building chatbots to embrace the emerging trend? If they do so, are chatbots going to displace human agents? No, as much as Hollywood sci-fi films would like us to believe that, we are nowhere near the brink of a robot takeover.

Two reasons chatbots won’t be replacing live agents any time soon:

1. Chatbot applications are only as good as the data they leverage

Chatbots rely on machine learning in order to function properly. Machine learning, in turn, relies on good training data to train the algorithms. Getting the right data fast is not a simple process. Organizations who are serious about building an effective chatbot need to have a sound long-term data strategy. After all, building the chatbot is not a “one-and-done” process but an iterative one where new learnings in the form of data are continuously used to improve the performance of the chatbot.

For the contact center, it would be best to start your chatbot off with very specific use cases that need low human-touch and where good quality training data exists. It’s important that such an implementation can either provide a faster turnaround time for the customer or improve agent productivity. Some good examples of chatbots designed for specific use cases that work well are in the banking industry where chatbots are meant for answering very targeted and straight-forward questions like “What is my account balance?” “When is my next payment due?”. You want to avoid casting a wide net and including more complicated service questions that can lead to a bad customer experience.

A strategy that enterprises can use to hone in on the best way to implement a chatbot is to start with their agents. Have live agents field customer queries first to collect data on the most targeted and straightforward questions.You can then use that data to train the chatbot effectively.

2. Importance of human interactions

Chatbots are not adept at understanding human emotions (like sarcasm) or recognizing high-stress situations. At the end of the day, whenever there is a need for emotional intelligence or critical thinking, humans are the heroes we need. A customer calling an insurance company to report that their dream home was destroyed by wildfires cannot be effectively managed by a chatbot. Such situations call for empathy and delicate handling, AKA a live agent.

Chatbot Magazine wrote a great article on perfecting bot-human handoff where chatbots front every incoming conversation to detect what a customer needs and then determine if and when an issue needs to be elevated to an agent. The key here is to analyze the conversation history and use that data to optimize. Essentially, we see chatbots (or AI) being really effective in a contact center when they are tasked with mundane tickets, leaving agents to focus on issues that require a human touch.

The future of contact centers is better and faster customer service through technology while retaining human connection. Chatbots, in congruence with live agents, can help get us there. If the capabilities of chatbots are combined with the emotional connection and elevated skillset of live agents, you will have the ultimate super agent.

At Sparkcentral we spend a lot of time both solving for and talking to top consumer brands at the forefront of the digital messaging experience. Message us if you would like us to share our insights or learn more about how Sparkcentral can help your agents.

Vadiraj Katti
Vadiraj Katti is a Senior Product Manager at Sparkcentral. His passion for putting customers first while solving hard product problems made him transition roles from Software Engineer to Business Analyst to Product Management. At Sparkcentral, he works on building product features that champion effortless customer service through digital messaging channels.


Please use comments to add value to the discussion. Maximum one link to an educational blog post or article. We will NOT PUBLISH brief comments like "good post," comments that mainly promote links, or comments with links to companies, products, or services.

Please enter your comment!
Please enter your name here