Building an iPhone App – Lessons Learned

The following slides are from the AIGA and IxDA event called “Digital Dialogues” on Wed Feb 30, 2012 at ISITE Design in Portland, OR.

This is a tour of the Grady Britton discovery, design, and development process while building the Choicelunch iPhone app in partnership with Knuckleheads. We’ve included some roadblocks to watch out for, and materials from the build including sketches, wireframes, focus group data, and screenshots of the final app.

The talk was only 30 minutes so that we could leave a lot of time for questions.

Notable questions

Q. Wait. It looks like you took a prototype into the focus group for testing. How can that be? If you didn’t know what features to test, how could you have built a prototype to test?

A. Choicelunch was already committed to building a mobile app, so we weren’t tasked with testing feasibility. They had a sense of the major features they’d need to port over from their website, however they didn’t have definitive data on the method by which customers would like to order on a mobile device. eg, by date or by entreé. We felt that by building a prototype to test this specific question, and by doing it at the end of the focus group after the feature sort, we could be doubly effective when we reviewed the video of the prototype tests.

Q. You ended up falling into a Lean UX style of working with Knuckleheads. (iterate on concept, design, and development of specific functions or screens during development) Without the big master UX spec document at the beginning of the development phase, how were you able to build a QA process?

A. We actually did end up delivering a big wireframe spec; it was just easier for my team to have it as a reference for design. I hear Knuckleheads chiming in here, “note: you can’t have enough annotations.” Point taken. Whenever Knuckleheads and my team would land on a final solution for a specific section of the mobile app, I would merge it back into the master spec doc and post it as the latest version. I think the final version was v11.6.

That said, the QA process only somewhat referenced that spec. It did help, especially for the functionality of the calendar specifically, but for the most part QA consisted of simply having as many friends (close to the project and not-so-close) use it in as many mobile environment situations we could think of. Thankfully, we only had to worry about customers of the Choicelunch program, so the concepts in the app were already familiar to the user and they were able to focus on the functionality of the mobile app.

Related Posts Plugin for WordPress, Blogger...