Designing the loyalty app for a 20+ location coffee franchise with no digital ordering

Deville Coffee is a Calgary franchise, running since 2008 and grown to more than 20 locations across Alberta and BC. I started there as a part-time barista, and served the same regulars the same drink every morning.

What they didn't have was an app. Ordering happened at the counter or through mobile web, while Starbucks, Tim Hortons and McDonald's all had one. That gap costs more than convenience. There is no way to repeat an order in two taps, no record of who the regulars actually are, and loyalty runs on paper cards that only work while you are standing at the till.

Mapping the regulars turned that gap into a person. David works in the building above a Deville and buys coffee twice a day. There is a Starbucks across the street with a rewards program, and the choice he makes every morning is not about the coffee.

An ordering app that turns Deville's daily regulars into loyalty members, by making a familiar order effortless to repeat.

80% of Deville's customers order the same thing every time. That single number decides most of the design: the fastest path through the app is not browsing a menu, it is getting back to the drink you already know you want.

There were no brand guidelines to inherit and no component library to build from, so every rule was one I had to set and then hold. That is the part of working alone that actually costs you: not the drawing, but having no one to check an edge case against.

What replaced a team was a method. I scored features on value against feasibility before building anything, so the decisions about what to cut were made once, in the open, rather than repeatedly and by instinct.

An informative product card lets someone choose with less work: the photo, the description, the price and the badge all sit together, so a decision takes one look rather than a tap into a detail page and a tap back out.

Placing an order is where most people hesitate. An out, a way to review or edit what you picked without losing your progress, is what turns that hesitation into a purchase. Testing found the cost of getting it wrong.

Quantity started at zero, so tapping Add to Bag added nothing. In V3 it starts at one: tap Add to Bag and one item is added, or if you set a quantity, that's what goes in.

The Iconify set wasn't cutting it, so I drew a custom one to match the branding. The label mattered as much as the icon: people told me browsing a menu feels like less pressure than placing an order, and Menu became the most used page in the app.

Pre-AI era project I'm still proud of. Designed and prototyped the old-fashioned way, manually setting up design systems, tokens, variables, and everything in between ;)

Having no team meant no one to talk micro-interactions or edge cases through with, so I built the substitutes: design mentors, books, videos, and a habit of writing the decision down before building it. What came out of that is the thing I kept. I got much better at making a call on my own and being able to say why.

If you have questions about this project, feel free to contact me.