08. May. 2019
Chris Rall

How to Use Decision Trees for more Efficient Customer Service

Managing internal information and projects has become increasingly complex. This is particularly challenging when handling multiple customer service channels, applications and agent turnover. Complexity makes consistency and accuracy difficult. So how can you simplify things without sacrificing KPIs or customer experience?

Agents face this challenge daily and need consider a multitude of factors, possible causes, and solutions to each request. With so much information available to customers online, inquiries are also becoming more complex. Customers can find the answers for simple issues themselves and tend to reach out for difficult ones. This leads to longer training times, agent insecurity, lower productivity and a decreasing First Contact Resolution rate.

What are Decision Trees for Customer Service?

Decision tree technology is a simple yet powerful way to assist agents handling complex requests. It is a simple If-then workflow which guides someone through a specific process, providing next steps and solutions based on each answer given. Consider how the answer to simple questions like “Do you use Android or iOS?” or “Are you a U.S. citizen” impact the next step.

decision trees for customer service

By building custom decision trees for your products, agents are actively guided through service inquiries step by step. This ensures:

  • Consistent performance throughout your contact center
  • No questions skipped or forgotten
  • Increased FCR & fewer call transfers
  • Happier agents

Using our simple visual creation tool, users can create a troubleshooting workflow with branches based on customer responses, all without writing a single line of code.

What is a guided dialogue?

Guided dialogues are the agent-facing front-end of a decision tree. Instead of the workflow view with branches for different customer responses, the agent clicks through a simple text and image-based dialogue. They are widely used for technical troubleshooting, for example trying to solve an issue with Microsoft Windows.

In the customer service world, it can be used for call scripting, step by step instructions or process manuals to replace long text-based documents.

Imagine calling your insurance company to change an existing policy. The agent needs to check several conditions to see whether this is possible including whether you are married, how long you have had the policy, whether coverage abroad is required and more. All these criteria influence the outcome and which, if any, solution is possible.

With a knowledge base that includes decision tree technology, all these questions are pre-defined, and agents can easily go through a request step by step. This ensures consistency among agents, whether experienced or brand new. Customers are satisfied and get the help they need, faster and without transfers. Decision trees will not only impact your first contact resolution rate, they will also provide the foundation for your future customer experience and NPS.

What is a guided dialogue?

But that’s not all! Let’s not forget Training

Decision trees not only help with direct customer service but also training. The question and response pattern can be used for your agent training to reduce on-boarding time by up to 80%. Even experienced agents benefit because they no longer need to know every possible complex request by heart. Having issues with agent turnover? Decision trees help you to make new hires productive from the very first day on the floor.

How to Approach Guided Dialogues & Decision Trees

Guided dialogues are the agent-friendly interface that decision trees underpin. To get started, we recommend the following:

  1. Identify your simplest inquiries (for example those already answered in a FAQ).
  2. Map them out and limit them to no more than 5 subsequent questions.
  3. Deploy them in your contact center and monitor performance and feedback.

Once the first few guided dialogues have been successfully implemented and used, it’s time to identify key areas where the need is most critical. These will be more complex issues where the solution is often dependent on multiple factors and where agents lose a lot of time troubleshooting.

Complex requests and service processes must consider multiple conditions and scenarios meaning they consist of different branches based on customer answers. This is where our graphical user interface really shines, enabling users to visualize the process without complex coding or logic.

Don’t Go Overboard

Not every customer request requires a decision tree! The goal is to save time on complex inquiries while increasing consistency and service quality. The time saved is then available for other requests which will immediately increase your overall efficiency and number of handled calls.

How to use decision trees for customer service

Decision trees build the future for your contact centers

Decision trees create clear, defined workflows that can be measured, evaluated and optimized. This is a welcome change for both managers and agents who benefit from greater transparency and consistency. With your whole team on the same page, it makes monitoring overall performance easier as well as identifying which areas work well and which need improvement.

Decision Trees for Self-Service, Chatbots & Alexa

Implementing consistent service workflows via decision trees will make agents happier, more efficient and your customers too. Moreover, the exact same decision trees can also be used as the basis for customer-facing self-service processes on your website. For many millennials this is not even the future, this is what they expect right now. In addition, existing decision trees can be combined with AI and natural language processing to form the brains of your chatbot or even Skills for voice assistants such as Alexa!

Finally, decision trees can be integrated into 3rd party applications such as Zendesk or Salesforce. With these integration capabilities, companies will get more out of their existing must-have applications.

Learn how to use decision trees with a knowledge base or request a demo.

There are no comments yet.