Vertex Chatbot

Overview

As part of our ongoing relationship with Vertex, we worked with their Cloud product team to reinvigorate their Help Center experience. From top to bottom, we systematically overhauled everything from their search capabilities to the navigation and page structure. While the experience has been very well received, there was still a need to address more complex questions. So, we partnered with Vertex to prototype, test and deliver a “support bot” to assist users.

The Environment

Enhancing the Support Experience

Sales and use tax can be a complicated space. While Vertex’s products have been streamlined to be more usable and intuitive, their customers’ needs vary depending on their particular situation. So, support personnel can be faced with challenging requests that require multi-tiered responses. However, 70-80% of these requests are often similar in nature. This was the perfect environment to automate a response with a chatbot solution.

Labs’ing Together

Prototyping & Testing with Vertex

We started this process by leaning on our years of labs experience and worked with Vertex to develop a prototyping and testing plan. First, we time-boxed our approach to develop a proof of concept to ensure we would not over-invest in the new technology. Then, we worked with Vertex’s support team to review common requests and developed various conversation threads that the bot could support in this first phase. We then tested the bot both internally with Vertex team members and externally with a few customers to validate its value.

The Roll Out

Launching the Vertex Chatbot

Once we validated the value of the new chatbot to Vertex customers, we refined the concept to handle seven key requests and added polish to the experience to make the bot more human and conversational. The addition of colloquial phrases, jump off points into search when the bot could not handle a task and slight wait times to account for the bot “typing” a response, made the experience feel more complete and helpful. Then, ongoing, we leveraged the Dialogflow (a platform we had learned through experimentation) to manage conversations and empty results from users.

The Future

Continuing to Innovate with Vertex

As we continue to develop the chatbot, we look at data to find new ways to improve the experience. What started as a proof of concept and test, has now turned into a full-featured roadmap and a solution we can extend to other ends of the Vertex product ecosystem. Our systematic approach to the build allowed Vertex to measure its investment into the new service bot, but also opened the door to learning about new experiences and technologies that can be leveraged beyond the Vertex Help Center.