3 minute read
It’s a Friday afternoon free of work and my brain is tickling. I’m feeling creative so I open my spark list (a list of ideas for projects I have been keeping for the last 5 years). After scrolling past some ideas that make me wonder whether my past self had hit their head that day I reach one titled ‘Lottery email’. Bingo. I wanted something I could build in a day and let fly to the world. The idea is simple. Every day one subscriber from the list gets to email everyone on the list. It’s my spin on giving people their 15 minutes in the spotlight, though you remain completely anonymous by default. I liked the idea because it’s playful, it’s human and it’s in no way business-y. I don’t want to spend my Friday afternoon building a tool, I’d like to build a toy or a curiosity that people will enjoy for its own sake rather than what it can do for them. So The Soapbox was born…
It’s more than just a curiosity to me. It’s a social experiment, it scratched my creative itch from inception to completion in an afternoon, and it continues to invite actual humans into my inbox daily.
Psychology has interested me for a long time. Seeing what people choose to post when they get the chance to say it to the world is fascinating. Speakers’ Corner in Hyde park tends to attract a certain crowd and it was an inspiration for the project, but The Soapbox has surprised me simply by the breadth of topics covered, though more on that later. The other interesting aspect to me is the virality of it. In order to gain more tickets to win the draw, users can refer others to the list. It’s a simple loop but it has proven very effective for encouraging referral growth and therefore giving winners a larger platform to speak to. I have never built something intentionally viral before so seeing how well it has worked (~44% of signups are referrals) has made me far more interested in it as a growth channel for other projects.
I’ve not vibe coded a completely greenfield project before so this was also an experiment in how well I could steer architecture and make decisions while not understanding every line of code or decision in minute detail as I might on other projects I’ve built. I’ve not been living under a rock so agents had already become part of my workflow, but always on projects with solid existing foundations. Being able to complete the whole thing in an afternoon really did make it more fun. In the past if I’d decided to build this I would have started enthused, spent a day or two on it and then been distracted by a newer shinier project. Not so when it’s ready to launch after 5 hours! Special shout out to Claude Design for spitting out something way better than I can do as a non-designer.
Spoiler: It’s Cloudflare all the way down.
I may be vibe coding but I still want to understand the codebase easily so I reached for SvelteKit which I’ve been working with and loving for years - Thank you Rich Harris et al. 🙏. The whole app is a single SvelteKit project built on Cloudflare Workers.
And crucially there is human moderation built in before every broadcast is sent to avoid spam, though so far only 1 person has tried to sell something.
For a fun Friday experiment the results continue to surprise me. 2 weeks after I hit launch I received an email from Cloudflare saying “Congrats on surpassing 10,000 pageviews on thesoapbox.email last month! This is the first time for your website.” which makes it the most popular thing I’ve built so far.
It’s been such a fun project to build and I’m very glad I saw it on my spark list. I’ll be continuing to let it grow organically as it’s doing and see where it heads next. Please do join the crowd for a daily dose of happiness.