Building an app without a reference point
As an iOS developer at ZOZO, Inc., Michael has worked on products like a 3D body measurement suit and a 3D foot scanning mat. These products are like nothing that have come before them, so building them requires innovation both from a technical and UX perspective.
Michael first came to Japan on a working holiday visa. While on the visa, he taught English and travelled around the Kansai region. He enjoyed himself so much that after returning to his home in New Zealand, he found himself wanting to go back to Japan once more, so he intensively studied Japanese, using the All Japanese All The Time method, while working as a developer.
A couple years later he returned to Japan, moving to a small rural town. As there was no work for a programmer locally, Michael turned online, and took international freelancing gigs that he found on places like AngelList and Upwork. But after a couple years he felt like something was missing, as he had spent so much time studying Japanese, but wasn’t getting to use it professionally, so he decided he’d have to give up country living and move to Tokyo if he were to work for a Japanese company.
Searching through postings, he found one for ZOZO, Inc., and decided to give their app a try. He said, “It was very useful and easy to use. I think the thing that made me want to apply there was they had shoes in my size. I have quite big feet, and so I can’t go to a shoe store here and find shoes in my size. But on ZOZOTOWN they had my size and so I bought a whole bunch of shoes, and I applied as well.”
I naturally grew into this role. It wasn’t like they created that role, and tried to find someone for it. It was like I was a natural fit for that sort of role, so they created a job around it, which was really nice.
Michael is now an iOS developer at ZOZO’s measurement team, which works on products to help users to find fashion items that match their unique body attributes. These include ZOZOSUIT 2, an accurate 3D body measurement suit, ZOZOMAT, a printed mat that allows customers to scan their feet easily in 3D at home, and ZOZOGLASS, a device that captures customers’ skin tone from the comfort of home. When he first joined ZOZO Inc., ZOZOSUIT hadn’t been launched yet, and the company was scrambling as their original approach of using stretch sensors, embedded capacitors that would change capacitance as they stretched, proved to not be cost effective. Instead of complicated sensors, they turned to a unique suit design, using fiducial markers, combined with three dimensional fitting algorithms.
These algorithms were developed by a firm which would later become ZOZO New Zealand Limited. That he would be working with people in his home country was a complete coincidence, but his background led him to become a bridge between the Japan and New Zealand teams. He said, “I naturally grew into this role where I was the glue that was holding a good deal of the project together, which was a bit of a shock for me, having just joined the company. It wasn’t like they created that role, and tried to find someone for it. It was like I was a natural fit for that sort of role, so they created a job around it, which was really nice.”
With brand new technology like ZOZOSUIT or ZOZOMAT, there are no real points of reference for creating an experience that works.
Michael enjoys the innovative challenges he faces with his work. He said, “We’re not building a shopping app for instance, where there are many examples out there that provide a reference for how to create that type of app. With brand new technology like ZOZOSUIT or ZOZOMAT, there are no real points of reference for creating an experience that works. There are also a whole bunch of technical limitations that must be conformed to, while at the same time providing an interface that won’t cause the user to get frustrated and give up. So there is a lot of room for creativity.”
An example of one of these challenges is that at the launch of the first ZOZOSUIT, the iPhone’s front camera didn’t have sufficient resolution to perform the measurement, so they had to use the back camera instead. He said, “The problem with the back camera is that the user can’t see the screen, so you can’t give them any sort of instructional feedback on the screen itself. Measurement feedback was made into voice instructions, and the design of the UI was centered around them, which made it an interesting problem to work on.”
Another challenge was with the ZOZOMAT, where users needed to hold their phone at an angle when doing the scanning to get enough depth information to create an accurate 3D model of a foot, but user testing showed they had trouble understanding the detailed instructions.
Michael thought he could simplify the scanning method, and got permission to develop a separate interface as a research project. It didn’t work out, as the processing speed wasn’t fast enough, but he was appreciative of the chance to try. He said, “That’s an example of how I was able to try something completely different, and with the presumption that it might not work at all, which was really nice. Being given that freedom as a developer makes you feel like you’re part of the product, and that you actually do have a lot of influence on a product that millions of people use.”
Ultimately, instead of a purely technical solution, they were able to refine the tutorial to the point where users no longer had issues with it. He said, “As a programmer, I thought it would be something that you should be able to solve with changes to the actual scanning process. It turned out that having a video tutorial of a person actually doing the movement that users could emulate made a big difference.”
These sort of challenges mean the measurement team has to work closely with the designers. He said, “There are many decisions that can’t just be left to the designers like with a web app. For instance, there might be certain algorithmic limitations or there may be technical opportunities that the designers aren’t aware of, so a lot of close collaboration is required.”
I was given a lot of time to research my talk, despite it not being directly related to product development.
While there are lots of opportunities to grow as an engineer through the work itself, ZOZO is also very supportive of independent learning. He said, “I gave a talk at iOSDC Japan, which is a Japanese iOS conference. I could talk about anything I wanted, it didn’t have to be pitching a product or even related to what I’d been working on. I was given a lot of time to research my talk, despite it not being directly related to product development. This was also the case when I did a talk at try! Swift, demonstrating ray tracing using Swift. Ray tracing is mostly unrelated to what I work on day-to-day, but once again, ZOZO. were very supportive, even though researching the topic and creating a presentation ate up a good deal of my development time.”
Michael’s team is very international, and most of them are able to speak English. I’ve never been belittled for not being able to understand something in Japanese. I’ve found the company to be very understanding that working as an engineer in a foreign environment can be challenging at the best of times.”
Michael sees it as a good stepping stone for those who want to use the language, but are just at the beginning of their journey. He said, “If you’re coming to Japan and you’re very interested in the culture and language, but you don’t think you can step straight into a role that uses Japanese only, I think it would be a great learning opportunity. To start out, you could communicate in English, but as you work, you’re learning just by being immersed in a Japanese environment. If you don’t want to learn Japanese you’re missing out, because it could almost be thought of as a free Japanese lesson, which is a win for everyone.”