Revolutionizing Customer Support with IBM Watson Assistant

Introduction

The objective of this tutorial is to demonstrate how quick and effortless it can be to develop a virtual assistant on IBM Cloud, once you have grasped the necessary techniques. Within a month, you can have a complete virtual assistant tailored to your specific industry and company. One of my favorite features is the natural language processing capabilities of the assistant, powered by scalable AI and machine learning technology, it continuously learns from customer conversations and improves without any involvement or cost. Additionally, it can use more than 60 languages.

Why Use Virtual Assistants?

Clients have reported several key benefits from utilizing virtual assistants, including improved efficiency with reduced support costs, allowing customer representatives to concentrate on more complex tasks while directing simpler queries to the virtual assistant. This, in turn, leads to improved customer satisfaction with quicker resolution of initial inquiries resulting in a more pleasant experience for the customers. The virtual assistant also provides a comprehensive end-to-end view of customer journeys, enabling access to valuable analytical data. This tutorial will explore the challenges currently faced by customer service teams, outlining how a personalized virtual assistant can address these challenges. At the end I provided an example and video of how it works.

To put the cost-saving benefits of virtual assistants into perspective, consider the scenario where 4 million weekly conversations are exclusively handled by human agents, which would cost the company around $8–12 million. By contrast, if Watson Assistant were utilized, the cost would be significantly lower, at around $900,000. This demonstrates how Watson Assistant can pay for itself and more. According to a Forrester TEI report, a 337% ROI can be achieved in less than six months on this investment. More than 1,000 reliable and trusted partners across a range of industries are currently utilizing this service.

Customer experience today is a struggle:

• 51% of agents without AI sat they spend most of their time on mundane tasks
• 3% of companies said they are able to act on all of the customer data they collect
• 78% of clients will back out of a purchase due to a poor customer experience

Current self-service for customers usually has the following challenges:

• Confusing navigation
• Stale and redundant content
• Inflexibility
• Not listening to feedback
• Forcing users down the wrong path

Key benefits of Virtual Assistant:

• Decrease in the number of support calls
• Increase traffic to the brand/company
• Decrease in response time
• Reduction in total support cost
• Increase in customer satisfaction

Now that we have discussed the benefits of Watson Assistant, let us shift our focus towards the process of setting up your first virtual assistant.

Watson Assistant Setup Guide

Go to IBM Cloud -> Search Watson Assistant in the top search bar and click enter. Select Create and fill out required information.

You will get a walkthrough of a tracker with first steps on the main page of establishing your assistant name, greeting message, etc. On left hand-side you have a column where you can click on actions to begin configuring your chat bot. Here is where you want to think about what features are important to your customers and how you will showcase it with different questions.

Here is an example of how you can create your own “actions” step by step with various options. You can make it so that the user dictates the path of the chat bot. In this example the user is applying for a loan. To begin you must tell the chat bot which phrases will bring up this action so you come up with words the customer will say by clicking on “customer starts with” on the top left.

In this example the user is applying for a loan. You can customize what the assistant will say when it recognizes what the client is looking for. I made it so that the virtual assistant will answer with “What type of loan do you want to apply for?”

After clicking on the “edit response” option you can specify the options that will appear to the user. In this case the assistant will want to know what type of loan the customer wants to apply for, and I provided the options.

In this step I used the conditions on the top saying that this path will only be taken if the user selects “Student Loan”. In this screen you can use the “$” function so that it treats the customers reply as a variable and brings onto this screen to certain text. At the end I use a “Confirmation” response to prompt the user if this information is correct (Yes/No).

\

I continue to create steps to get more information from the user such as what degree they have and what state they are located in. I also include a step in which if the user clicks “No” to the confirmation question then this action will end since it is not what they are looking for.

At the end of this I use the “$” function to get all the answers that the user chose all in one screen and ask for confirmation that it is all correct.

Summary

We discussed the advantages of utilizing virtual assistants for customer service. These included enhancements to efficiency and reductions in support costs, as well as improvements in customer satisfaction and access to valuable analytical data. It is worth noting that the creation of a functional virtual assistant is a straightforward process, as demonstrated in the tutorial. Implementation timelines are short, with a co-creation workshop taking 1–2 weeks to tailor the virtual assistant to your company’s specific needs, followed by a 3–4 week period to build and deploy the assistant onto your website. One added benefit of utilizing Watson Assistant over other AI services, such as Chat GPT, is that Watson exclusively employs your company’s data, leading to greater accuracy, control, and governance, with the ability to verify its responses.

Watson Assistant Demo

This is an example of what the assistant would look like. There are many options in the homepage and on the top of this page where you can customize the web chat, change the background, or incorporate it into a website.

Summary

In summary, the demonstration centered around the capabilities of IBM Watson Assistant in addressing key challenges that current chat bots currently face, which include a basic chatbot interface, unanswered customer queries, unavailability of live agents, and a lack of clarifying questions. Moreover, the presentation highlighted the key benefits that customers and businesses can expect to achieve through the implementation of this virtual assistant solution. In the video I showcase the practical applications of the Watson Assistant in real-life scenarios. Feel free to comment or reach out for any questions!