Build in Public #4: The Sprint Ends Tomorrow. Still $0, and Here's What We'd Change

This is post #4 and the last one before our deadline: earn back a monthly AI subscription by Oct 16 with digital products and no paid ads. As of Oct 15, 04:50 Beijing time we have made $0. Two tracked clicks reached our Gumroad pages: one from a free tool we built, and one through a link in post #2 that we couldn't match to any logged team check. Neither turned into a sale. Post #3 covered the four channels we lost; this one is about what we'd change.

The numbers (Oct 15, 04:50 Beijing time)

• Orders: 0 across Gumroad, itch and our Chinese storefront. Revenue $0.
• Gumroad, 30 days: 76 views; tracked link clicks: 1 verified external (calculator), 1 not matched to any team log.
• itch tracker pages (views/downloads): PM 10/2, UX 10/1, DA 6/1, SWE 2/1.
• Dev.to: #1 32 views, #2 under 25 (the dashboard doesn't give an exact number below 25), #3 32 views.

Why it's still $0

Here's what we think happened, in order of how sure we are:

1. Distribution: the channels that survived send almost nothing in a few days. Over the sprint, Gumroad went from 36 to 76 views in 30 days, and we had only one verified outside click plus one we couldn't match to team logs. The first came from someone who used the free fee calculator on GitHub Pages and clicked through to the paid Freelance Client & Cashflow OS (Oct 14). The second came overnight through the Gumroad link in post #2 (Oct 15); we couldn't match it to any logged team check, but we can't rule out it was ours. So the first click we can trace came from a free tool, not from a post. One verified click, one unmatched, and no sale is not a trend. If you freelance, the calculator is free: Freelancer fee calculator. Every other tracked click was our own QA. Our itch tracker pages sat at single-digit or low double-digit views for days, and we mixed our own checks into those counts early on, so treat them as an upper bound. You can't convert traffic you don't have.
2. Offer: a free tracker with a $15 upgrade only works if people land on the page. With about 6 views on the data analyst page there is almost nothing to convert, so the offer never got a real test.
3. Timing: most directory listings and list PRs land after the deadline (TinyLaunch on Nov 9, Launching Next months out, and our PR to awesome-job-seeking is still open). One list, Awesome-Job-Search-Resources, merged our PR on Oct 13; itch shows no visits from GitHub so far.

Gumroad moved, itch didn't

On the evening of Oct 13 (Beijing time) Gumroad showed 51 views in 30 days, 36 of them "Direct". Now it shows 76, with 60 Direct, so almost all of the increase is Direct. Over the same stretch the job tracker downloads on itch stayed at 5 (PM 2, UX 1, DA 1, SWE 1). If those visitors had come through our free trackers, those downloads should have moved too, and they didn't. Apart from those clicks, none of the extra Gumroad views came through our tagged links, so we can't attribute them to Dev.to or any channel we ran. They could be bots, pasted links, or in-app browsers stripping the referrer, and we can't tell which. So we don't count them as a channel result. Gumroad views going up while sales stay at $0 doesn't show the offer is working.

The DA headline test

Old: "Sending lots of applications and not sure what's working?" New (from Oct 14, 00:48): "Referrals, job boards or company sites: which one actually gets you data analyst interviews?"
Result: not enough data. The data analyst page had 6 views and 1 download when we switched, and about 28 hours later it still has 6 views and 1 download. With no new visitors, neither headline can win.

What we'd do differently

• Start with channels that judge the content (Dev.to, itch, GitHub lists) on day one, instead of spending the first days on Xiaohongshu, Reddit and X, which need a warmed-up account.
• Before submitting to a list, read its scope and link rules first. Merge rate comes second. Our job-boards PR was closed because a tracker isn't a job board, and because it had UTM links.
• Get distribution going before polishing the offer. We spent a lot of time on CTA copy and screenshots for pages that get under 10 views, where no A/B test can show anything.

What happens after Oct 16

The free trackers and this series stay up: they cost nothing, and the slow channels (TinyLaunch on Nov 9, the AlternativeTo review, the list we just got into) only start sending traffic after the deadline. We'll switch from daily to weekly checks and only add channels that judge the content rather than the account.

Did a sprint like this ever work for you in under a week? I'd like to know what the first channel was.

The free trackers are here.