Intuit Tinder
Improving industry-leading software solutions for marketers and sales enablers by simplifying and optimizing industry-standard tasks.
ROLE
UX Designer
UX Researcher
TEAM
Kris Nagoda - Product Manager
DATES
Apr 2018 - June 2018
Problem
TurboTax, a leading tax tool, is used by millions of consumers and professionals annually. It offers a comprehensive feature called Responsive Naturalized Answers (RNA), which tracks filed forms, identifies incomplete fields, and predicts user questions based on this data. With significant changes for the 2018 tax season, Intuit had to quickly scale resources to support and update the RNA platform within a tight deadline.
RNA in TurboTax
When RNA was initially conceived it was quickly implemented without a focus on scalability or ease of use. The initial workflow took 6-8 weeks for a product owner, writer, program manager, back-end developer, and front-end developer to get from concept to published product for each question and answer.
RNA v1 workflow
Hypothesis
I thought that the manual creation and implementation of RNA led to inconsistencies in the models and increased the potential for logic errors as tax laws changed. To address this, I believed focusing on reproducible workflows and easy validation would ensure consistent logic and reduce errors in the developers' code implementation.
Process
I first met with the product manager to understand the current workflow and recent solutions. As the sole user, they had created a Virtual Machine (VM) on their laptop to streamline the handoff of Visio files into a rough code base for developers. While this reduced time by weeks, it limited team expansion, as each user needed a custom VM and training to use it. (This also failed the bus test).
RNA v2 workflow
The product manager walked me through the current implementation which included mapping our the logic by hand in Visio before exporting it through a custom software solution that would hand off rough code to his developer.
Mapped Visio flow
Java code base
The product manager mentioned that much of the developers' time was spent debugging new logic based on existing code, something he hoped could be solved with cleaner output from their solutions. Since our internal team had recently launched a WordPress integration, we saw an opportunity to create consistent code output. We then mapped out key use cases for his new team members: creating a new topic ID, editing and testing a topic ID graph, and editing and testing content changes.
Create new topic ID
Edit and test a graph
Edit and test content
After testing and validating the concept, we discussed with our Wordpress team and found quite quickly that seamless integration of the tools would not be possible. We did find we could quite easily implement an iframe solution that would allow us to quickly implement a solution given the short time-line available.
Tinder in iframe
Highlights
The tools team was able to have an immediate impact on the RNA team's workflow. Implementation of the design resulted in a time reduction of 1 week to 5-30 minutes to implement answers and questions. Not only was time significantly reduced, but the publishing process became much more consistent and Intuit was able to save $3 million on this project.
Learnings
This project was formative for my career as it taught me to engage with a variety of subject matter experts early and often when exploring solutions. Had I not been proactive about understanding potential solutions we would not have been able to identify clear valuable long term roadmaps that made immediate impacts for current users while also having a clear path for how the product would scale as resources became available.