In November 2020, I conducted a survey of international developers living in Japan to better paint a picture of what the typical foreign developer life is like. The results are now live, please check them out!
With 362 people responding, I had about 50% more responses than 2019’s survey. While in absolute terms, that isn’t a very large turnout, given the very niche audience I’m targetting, it’s quite a respectable sample.
A team effort
Given that so many people took their time to help me out with this, I wanted to repay their efforts by coming out with something more than just a simple summary of them like I’d done in previous years. On my own, I realized I could only do so much, so I enlisted the help of others in it.
Tom Coombs, who I originally knew from his role as the organizer of one of my favourite events, UX Talk Tokyo, helped with the data analysis and visualization. He’s also written an article about his experience building the charts with d3.
Mathieu Mayer, who had previously helped me with TokyoDev’s redesign, helped with giving the results a visual polish.
Many others helped with previewing the results before they went live.
A little outreach goes a long way
This year’s survey asked about gender. Forty minutes after launching the survey to my audience, I had 41 responses, but only two from women. Not only was that a bad turnout percentage-wise, in absolute terms it was looking so miniscule that I wasn’t going to be able to make any meaningful correlations with gender.
I turned to Tutti Quintella, a group manager at Mercari and director of Women Who Code Tokyo about how to better reach a diverse audience. She suggested I explicitly put out the call to women to answer and that I reach out to women-heavy communities, and also offered to share it through Women Who Code Tokyo.
With just putting in a little more effort, I ended up with 12% of respondents indicating they were women or non-binary, more than two times what I had got before I did any explicit outreach.
Start planning earlier
2020 was an exceptional year, and I think we all had a lot on our minds. It wasn’t until late October I realized that if I was going to conduct another survey this year, I had to get moving. This ended up rushing things a bit more than I would have liked, and didn’t give me as much of a chance to do community outreach before I actually launched it.
The next survey will launch January 2022
One of the challenges with doing a survey like this is that the value of the responses degrade over time, so there’s a tension between quickly putting something out, versus spending the time it takes to get something polished.
Conducting the survey in November meant that I was still working on the results over the New Year’s break, a time when it seems like it should be able to get some work done, but you never make as much progress as you anticipate. So for the next survey, rather than launching it at the end of this year, I’ll do it at the beginning of next year instead.