Discovery

My Role

I was brought onto the SwiftSense chatbot project to design chatbot that captured the SwiftSense's brand voice while helping users navigate the flagship site.

Brand Overview

SwiftSense is a mid-size SaaS data visualization platform that caters to small business owners inexperienced with data analytics. SwiftSense’s slogan, “Analytics for All,” reflects the brand's commitment to data democratization through accessible UI, hassle-free integration, and affordable price plans.

Brand Voice

Friendly, casual, and somewhat goofy.

User Overview

The chatbot's primary user base is small to medium sized business owners, IT teams, and marketers navigating the flagship SwiftSense site. These professionals possess varying levels of data literacy

Project Limitations

- Due to limited NLP data, user input buttons would prioritized over capture inputs

- Limited user research and testing analytics

- Chatbot must be designed within VoiceFlow

Define

Brand Goal

Shareholders expressed the primary goal of the chatbot was to reduce call center burden. By leveraging a chatbot to resolve simple user queries, call centers could focus efforts toward resolving complicated issues.

User Goals

Based on previously discovered user interview research, a typical SwiftSense.com user visits the site to achieve one or more of the following:

- Connect to a sales agent

- Connect to technical support

- Learn more about the SwiftSense's product and value proposition.

Design (Flows)

Based on the brand goals, user research, and user goals, I decided to design three primary flows: a sales flow, a technical support flow, and information flow. The sales and technical support flows would generate respective support tickets, while the information flow would provide information on SwiftSense’s products and share links to the SwiftSense blog.

In addition, these primary flows would be supported by four contextual flows: a hello flow, a catch-all flow, a human-hand off flow, and a goodbye flow.

Each flow would be designed to seamlessly connect with one another. The result would be a chatbot conversation that would end with either the user achieving their goal or, as a last resort, connect them to a human service agent to resolve their issue.

Design (Dialogue)

"Swiftie" The SwiftSense Chatbot

I chose to give the SwiftSense chatbot a simple, friendly name to reflect the brand's voice and commitment to accessibility. Swiftie wants to make small business owners feel welcome, but understands that they have visited the site for a specific goal, not to chat with Swiftie. They may not know how to navigate the site or where to start, so Swiftie guides users with gentle efficiency.

Polite and cooperative. Swiftie wants to work with users rather than just tell them what to do. They are always polite and use language like “please” and cooperative phrases like “let’s” and “we.”

Extroverted, but not overwhelming. Swiftie is eager and excited to chat with users but writes simple, short responses whenever possible.

Goofy but task-oriented. Swiftie has a goofy sense of humor and loves to crack contextual jokes, but only when appropriate. It tones down its humor when a user may be frustrated, such as when they need to contact technical support or experience a no-match.