If you are building a chatbot, learn all about chatbot deflection, why they are important, and measuring deflection rates that matter.
Chatbot deflection is also often referred to as call deflection. Chatbot deflection workflows allow organizations to reduce call volumes effectively and efficiently while also potentially providing a better customer experience.
In addition, chatbot deflection also reduces the workload of human customer service agents to refocus their efforts on more complex problems without being overloaded with smaller issues. Used in combination with automation and emerging AI, chatbot deflection is a promising solution for organizations looking to reduce call volume and staff shortages.
Gartner recently suggested that 40% of current live volume of calls could be diverted and solved with self-service channels, and businesses are currently starting to only recognize this opportunity for chatbot deflection. Combined with other call deflection strategies targeted on reducing call center volume, chatbots can help decrease human staff workload and instead refocus their efforts on more complex problems.
All About Chatbot Deflection
Chatbot deflection represents an event when a chatbot and a customer service team work the same hours, and the chatbot initiates a chat interaction that attempts to solve a customer’s problem. A deflection is deemed successful when the chatbot’s interaction has deflected the customer way from interacting with a human customer service representative.
High call volume can be stressful for customer service teams, especially during peak periods, because customers can become irritatingly frustrated when they have long wait times. This can decrease customer satisfaction scores, thus impacting brand reputation, and driving traffic to other competitors. The focus on reducing inbound call volume is a way to keep a pulse on the customer experience, and this is where chatbot deflection can help.
Creating experiences that facilitate an easy interaction with a business is a great way to build a customer’s knowledge about products and services as well as helping reduce call volume.
Proactive Chatbot Deflection
A proactive chatbot deflection allows the bot to reach out to them to communicate messaging and provide service updates. This allows customers to self-serve to resolve inquiries on the fly preemptively without having to wait for a human customer service agent or salesperson. Proactive chatbot deflection scenarios can include handoff processes to a human. These can be seen as preventative measures to reducing incoming calls to customer service centers, saving human staff unnecessary tasks and calls that would have arose without the proactive chatbot deflection.
Reactive Chatbot Deflection
A reactive chatbot deflection allows customers to reach out to the business or organization and choose an alternative channel of communication. This type of deflection drives the the interaction on the channel of the customer’s choice.
Calculating a Chatbot Deflection Rate
A chatbot deflection rate can be calculating by totaling the number of chat interactions during customer service hours, dividing it into the number of interactions that are transferred to a human customer service representative or agent, subtract that result from one, and multiply it by 100 to get a chatbot deflection percentage rate.
Workflow of Chatbots Deflecting from Human Agents During Office Hours
Generally, chatbots can automate a request from a customer/user or deflect the conversation from using expensive human capital resources. A chatbot implemented to help deflect from these resources will be positioned to interact with the customer at the first point of contact with the customer. If the chatbot gets confused, it will be presented with a disambiguation dialogue, potentially creating an open ticket to be handled by a human agent, schedule a call back time, or provide an email with more information to the customer.
Workflow of Chatbots Deflecting with No Human Customer Service Representative Agent as Backup
When a chatbot doesn’t know an answer during an interaction when it is outside of office hours, the experience is slightly different. Instead of handing over the conversation to a live human agent, the general conversation dialogue design includes an empathy message apologizing to the customer, then providing a disambiguation dialogue that provides other options including allowing a customer to provide details for a callback from a human agent, providing self-service knowledge articles, and providing a call back time-period when live agents are available.
Frequently Asked Questions
What’s the difference between a chatbot automation rate versus a chatbot deflection rate?
A chatbot automation rate is derived by taking all conversations the chatbot had into the interactions the chatbot was unable to solve, subtract that result from one, and multiplying that number by 100 for the percentage rate of chatbot automated interactions. A chatbot deflection rate is specific to the total number of conversations during customer service hours where a live agent is available and that were deflected.
How do you measure if a chatbot is successful?
When a chatbot is initially developed, a value proposition should have been defined as a baseline for what the chatbot was developed. Based on that value proposition, the chatbot return on investment should have been defined. Chatbot and conversation KPIs can be stablished in order to measure more discrete metrics to better iterate and develop features and content to help support the initial value proposition and return on investment.
Summary About Chatbot Deflection
Chatbot deflection workflows allow organizations to better serve customers in the information they are seeking (and potentially even providing preventative measures). Chatbot deflection workflows can help solve knowledge gaps for customers as well as provide a seamless and frictionless experience when handing off to a customer service agent. Having a chatbot deflection mechanism in place has a parallel benefit of reducing the call volume to customer service human agents, reducing the cost of human capital and/or shifting that work to more complex tasks.
Chatbot dialogue, conversation, and interaction metrics can help provide more insight to traditional KPIs used for customer satisfaction. These metrics can help correlate or provide inference data in what types of information customers may be seeking.
Other Chatbot Metric Posts You Might Like
If you are interested in other analytics and metrics around chatbots and how they can help businesses better understand their customers, check out the posts below:
- Measuring the Effectiveness of Chatbots
- Chatbot Limitations to Be Aware Of
- Chatbot Trends and Statistics
- Chatbot Return on Investment
- Chatbot Data Architecture