Product design + developmentMarch 2020 — February 2023

Swiftly.co.nz

During lockdown, 1,200+ New Zealand stores sold over NZ$700,000 through Swiftly.

Swiftly.co.nz homepage showing the e-commerce platform

Role

Designer, Frontend Developer

Period

March 2020 → February 2023

Categories

WebUX ResearchProduct DesignWeb DevelopmentDesign SystemsB2B2CeCommerceWeb Accessibility

About the project

I co-founded Swiftly with two friends and worked on the product design and frontend. Small New Zealand shops could set up a store in minutes, take orders, track stock and send invoices from the same app.

Challenge

Starting point
Under New Zealand’s April 2020 Alert Level 3 rules, most shops had to keep customers outside. They could take orders online or by phone and offer contactless pickup or delivery. Only half the businesses we surveyed could already sell online.

Response

My work
At the Hack the Crisis hackathon, I led UX research with shop owners. Our team launched a working storefront with inventory management and invoicing in 48 hours.
Outcome
More than 1,200 local stores used Swiftly to sell over NZ$700,000 in goods during lockdown.

Impact

48 hours

From idea to launch

1,200+

Local businesses on Swiftly

$700K+

NZD in earnings

17,000+

Orders fulfilled

168,000+

Unique users

1.2M+

Pageviews served

Initial steps

We started by choosing a hypothesis: "How hard is it to sell online?"

The entire team performed quick & dirty ethnographic research remotely with small businesses, as well as usability testing on competing products. It turns out that only half of our surveyed businesses were able to sell online, and many small businesses struggled with incumbent offerings.

One of the participants took hours to create a Shopify store, and many participants ended up either failing to complete the test, or quitting in frustration. To avoid contamination, we told them nothing about our hypothesis, and that we were just interested in how they ran their businesses during COVID.

As a team, we then synthesized the research into user needs, which were then mapped into functional requirements, user flows and a sitemap. We had to account for how customers could receive their orders under the lockdown rules. Read the April 2020 retail guidance.

We then mapped out the data into a simple schema, and split the work responsibilities between functionality and markup so that development could begin as barebones functionality while I fleshed out the rest of the user flows and design.

Research and documentation artifacts showing user needs synthesisUser flow documentation and sitemap

Lo-fi prototyping

To keep things in sync as a remote team, we livestreamed the designs as they were being mocked and collaboratively proposed designs to each other as the wireframes developed. Lo-fi designs that only demonstrated the happy path as well as expected behavior were explored, and subsequently developed in parallel as the high fidelity designs were put together.

Lo-fi wireframes showing initial storefront layout explorationsLo-fi wireframes showing checkout and inventory flows

Hi-fi prototyping

By the time the functionality was built with barely any CSS markup, the high fidelity designs were done. Given that we only had 48 hours to ship, I made extensive use of Uber's Base Web component library to save time and not have to deal with tricky components like complex inputs. After completing the hi-fi designs, I joined the rest of the team on development by writing frontend code and polishing bits of the user flows that needed attention.

High fidelity designs showing the polished storefront interfaceHigh fidelity designs showing product listing and cart

Frontend development

Most of my remaining time was spent writing react components, and cleaning up how components behaved responsively. I usually pair a tool called Solis with Visual Studio Code to review responsive behavior in real time.

Responsive behavior testing across multiple viewport sizesDevelopment environment showing component implementation

Post-launch

An hour after we launched Swiftly, we had ten organic signups. After a week, we had about 200. To date, we helped 1,200+ stores earn over $700,000 in revenue. The service was shut down in February 2023 as it served its purpose.

Swiftly analytics dashboard showing growth metrics
Final Swiftly platform showing completed storefront design