Groupe Dynamite · 2024 · Senior Product Designer · 4 weeks · Shipped · Qualitative
A new payment API in four weeks, no time for user testing, and most leaders expected failure. I refined the design daily with the PM and developers, iterated every screen inside its user flow, and split the release in two to protect conversion. Handed off on time, across two brands.
The CTO gave my product team an urgent project: refactor the payment gateway of both native apps, Dynamite and Garage, around a new third-party API. Four weeks to prepare development, no time for user testing, and most leaders expected it to fail.
I asked for daily refinements with the PM and developers, trading usability against back-end limits screen by screen. Baymard research, usability heuristics and the web team's experience stood in for the tests we couldn't run.
I started from mid-fidelity wireframes and stakeholder feedback gathered a year earlier. Two days in, it was clear we wouldn't make it. I went straight to high fidelity and put the screens inside the user flows so everyone could follow to protect the date.
Release 1 integrated the API without hurting usability or conversion. The trade-off was an inconsistent visual language across the flow. Release 2 filled the gaps. Side-by-side rows of the current app, Release 1 and Release 2 let non-product stakeholders see progress instead of remembering it.
Proactive work helped facilitate visual design. A master button, a master product card, one type system per brand, every state specced for developers. They became the first official tokens and components of the apps' design library.
I was hired to lead the complete redesign of Groupe Dynamite's two native apps. Groupe Dynamite is a Canadian fashion retailer with over 300 stores and two brands, Garage and Dynamite. I did it alone, and the redesign quickly showed problems bigger than the screens: no clear leader at first, siloed teams, no product development process, and a split between short-term conversion and long-term loyalty. App design was treated as surface work.
With no product team, I did the groundwork myself. I benchmarked About You, Zara, Adidas, Mango and Ssense, audited the apps against Baymard research and usability heuristics, and ran user interviews. I drew more than 300 exploratory wireframes and wrote 40+ design tactics grounded in behavioural science. I presented the work more than 40 times and tracked every piece of feedback in a table I revisited at each iteration. The new Home and universal UI landed with the CEO, who is notoriously hard to impress.
It's tempting to rush design work. I chose alignment meetings, transparent communication and strategy first, and trust became the key result. Stakeholders started to see how design decisions tied to business goals. That trust is what the next project ran on.
Then the CTO assigned an urgent project to my product team: a complete refactor of the payment gateway, because of a new third-party API integration. We had four weeks to prepare development. A back-end-heavy, high-risk change, with no time for user testing. Most leaders expected failure. I embraced the challenge.
This project wasn't driven by a product or user problem. Through ongoing collaboration, we found the issues the API integration caused, and I kept usability a priority until the PM, the developers, key stakeholders and I reached consensus.
I advised the team to hold daily refinements to align on the trade-offs between usability and back-end constraints. Without user testing, every detail had to be discussed to avoid breaking the experience and causing a conversion disaster. The strategy was simple: think like the consumer, apply usability heuristics, lean on past research and the Baymard Institute, fix known UX issues, and use the web team's expertise.
A year earlier, low-fidelity wireframes had confused stakeholders, so I started at mid fidelity. I drew two checkout shells: one with a sticky total that shows the points you'll earn, one that makes the order easy to double-check. Address and shipping got clear states, with a green check when information is saved. The Order Placed screen got three options. That screen is the peak of the experience, and the right moment to celebrate rewards or ask for something back.
I brought ideas worth stealing: Save for later from Lululemon and Amazon, swipe to remove, and my own sketch of a "$15 away from free shipping" bar. After talking it through with the PM and developers, we agreed it would exceed the project's scope.
Leaning on past work was a good strategy, but two days in, it was clear we wouldn't make it. High-fidelity wireframes became essential. Reviews went faster, confusion dropped, and developers juggling constant context switches could work straight from them.
One technique I love is putting the wireframes or mockups inside the user flows. Collaborators can follow along, take ownership and bring their best ideas. Stakeholders got a quick recap of what had changed since their last look, so alignment stayed fast and linear and we could move on to the next task.
The API's payment bottom sheet came almost as-is out of the box, and it sat at the centre of the refactor. We tried redesigning it, but we had to keep changes minimal to meet the deadline. Developer spikes kept reopening the payment step, which took several rounds of design over two weeks.
We covered returning users with past orders, error handling, and every payment path: Visa, PayPal, Apple Pay, Klarna, gift cards, loyalty points and their combinations. We also covered address validation, billing addresses and the shopping bag.
Every screen was aligned with the brand teams and specced to the pixel, with auto-layout throughout: spacing, both brands' type systems, a master button, a master product card with eleven parts. The Business Analyst used my documentation to fill Jira. Those tokens and components became the foundation of the design library for every app refactor that followed.
To minimise risk, we committed to two releases. The first integrated the API without compromising usability or conversion; the second finished the experience. Non-product stakeholders could compare rows of the current app and both releases side by side, so tracking progress wasn't a memory exercise. Every use-case decision went back to the VP of Product and the apps' business owners.
Over 100 wireframes, 80+ mockups and 25+ user flows were refined, reviewed and handed off in four weeks. The year before made that pace possible: the alignment, the feedback tables, the trust. Leadership isn't about authority. It's influence, earned through consistent, transparent communication. What I still don't know is whether conversion moved. We set out not to break it, and nothing I have says we did.
Flows, UI, specs, design library. With the CTO, PM, Business Analyst and developers; decisions reviewed with the VP of Product and business owners.
No user testing. No conversion data recorded after launch. I left before seeing the impact.
The goal was to maintain or improve conversion, and I have no data either way, so the claim is that we didn't break it.
Next case: The table at 45 degrees