Intuit TurboTax Canada
role
Marketing & UX Design Research Intern
team
2 Designers, 1 Product Manager
skills & tools
In-Depth Interviews
Usability Testing
Figma
duration
1 month
(Oct. 2025-Nov. 2025)
overview
TurboTax Canada was about to ship a feature that let filers skip manual expense entry entirely. But would Self-Employed users trust it enough to actually use it?
I partnered with a senior product designer to run moderated usability sessions with nine returning self-employed filers, testing the new CSV upload feature against their real spreadsheets instead of sample data. We wanted to know if it felt trustworthy enough to replace the manual process they already relied on.
We weren't just testing whether the upload flow worked. We were testing whether it felt safe to hand a year of expenses over to it.
What we found wasn't a usability problem. It was a trust problem.
problem
Filers weren't being asked to adopt a feature. They were being asked to trust a black box with their taxes.
Every self-employed filer has their own system. A spreadsheet shaped by years of trial and error that only makes sense to them. However, the new upload feature asked them to hand that structure over to an algorithm.
8 of 9 participants' spreadsheets would have failed to upload without manual reformatting first.
My research objective was to understand these habits firsthand, and test whether the new CSV upload feature, which was still in MVP, was both usable and worth trusting.
process
We were not testing whether the flow simply worked. We were looking for the moment someone stopped trusting it.
Before sessions began, I built and maintained the clickable prototype in Figma using the team's existing component library. It was my first hands-on experience making product design decisions inside an established system, and that work also surfaced components that needed updating, which I flagged and fixed ahead of testing.
I ran and moderated 9 usability sessions via Zoom, watching real self-employed filers test the flow of uploading their actual expense spreadsheets. I watched for the moments they hesitated, second-guessed a category, or scrolled back up to double-check something the system already confirmed.
5 of 9 participants said they'd enter expenses manually or test with just a sample first. They didn't trust the upload enough to hand over their real data.
Across sessions, four consistent patterns emerged: low trust in the upload's accuracy, confusing asset classification, hesitation around tax incentives, and a category list that didn't match how people think. I worked with the product designer to translate each into a concrete recommendation for the team.
key findings
Four patterns surfaced across sessions, and each one pointed to a concrete fix.
I watched nearly all nine participants hit the same four walls. Here's what each one turned into:
Spreadsheets wouldn't upload cleanly
I recommended a downloadable expense template so filers' data matches what the system expects before they even upload.
Assets people couldn't confidently classify
We moved to auto-assigned asset classes instead of leaving it up to guesswork.
Incentives people didn't trust
I recommended tightening the guidance so it's clear why someone qualifies.
A category list that didn't match how people think
We reordered it around actual filer behavior instead of internal logic.
None of these were major feature changes. Each one was aimed at the same thing: rebuilding trust at the exact moment it had broken down.
user snapshots
I built an individual snapshot for each of the nine participants: their pain points, their confidence level, and what it would take for them to trust the feature enough to use it again next year. These kept the team grounded in real filers instead of edge cases.
reflections
Clarity is sometimes the entire redesign.
None of it required rebuilding the upload feature. Every fix was small and specific, the kind of thing that's easy to overlook until you actually watch someone hesitate over it. That's what stuck with me most: clarity isn't a nice-to-have on top of a feature, sometimes it's the entire redesign.
Working alongside a senior designer showed me how much weight small decisions, like a confirmation message or a categorization label, carry in whether someone trusts a product enough to keep using it.
On a more personal note, this was the first major UX research project I got to own and lead, and it gave me real hands-on design experience too. It kickstarted and confirmed how much I want to keep pursuing product design. I loved getting to know the different business owners I spoke with, hearing what their day to day actually looked like, and it reminded me how much tech can genuinely support people when it's done right. That's the kind of product I want to keep building/designing for!
my team + memories
An unforgettable 8 months <3
So glad to have experienced living in Toronto for this internship! I am incredibly grateful for memories and friendships made during my time here :)
