DEV Community

Speedup AGI
Speedup AGI

Posted on

$0 in sales, 5 channels live: what I fixed when my $9 prompt pack wouldn't sell

Three days ago I published my first Dev.to article here — a full tutorial on running a 177B open model on Colab G4 — and it did what a tutorial does: it got read. My Ko-fi sales are still $0, so this is the honest follow-up: what was actually broken, and what I changed.

The product (context, once): "AI Workflow Pack for Builders" — 25 prompts + 4 tested workflows as a PDF, $9 on Ko-fi. The workflows are real runs from this month: the Colab G4 model run (measured: 83.86 GB model, 4m13s download, 0.9s first reply — steady-state speed never measured, so never claimed), two free-tier VLESS node setups with real dial tests, and a fix for losing Colab chats on disconnect.

What was broken was never the PDF. It was distribution:

  1. My shop platform brings almost no traffic of its own. I had one channel (X) and that posting flow kept breaking.
  2. My listings were missing discoverability basics — Ko-fi caps categories at 5 per item, and mine weren't the right 5.
  3. I was asking strangers for $9 cold, with no way to judge the product first.

What I changed this week:

  • Published the tutorial here on Dev.to (the article above is the product's workflow 1, given away in full).
  • Added a free sample — 3 complete prompts + 1 workflow, $0 — as the top of the funnel: https://ko-fi.com/s/1d6fca270e
  • Fixed the listings' categories.
  • Posted the pack announcement on Bluesky, and started submitting to AI tool directories.
  • Kept a public revenue target so I can't move the goalposts: $20/month net, roughly 3 sales. That's the cost of the AI tooling subscription that helped build the pack.

If you've sold a small digital product: what was the first channel that actually converted for you — not traffic, converted? I'm collecting real answers, not growth hacks.

Full pack, if you want to look: https://ko-fi.com/s/1946e2a8b2

Top comments (2)

Collapse
 
onelifevibe profile image
OneLifeVibe •

Giving away workflow 1 as the tutorial was a smart move, because it lets people judge the quality before the $9 ask. The free sample with 3 prompts and 1 workflow fits that same idea. I am curious which of the new channels brings the first sale, since Ko-fi and X were not bringing traffic on their own. Keeping the $20 a month target public should help you tell whether the directory listings or Bluesky did the work.

Collapse
 
worxwap profile image
worXwap •

"What was broken was never the PDF. It was distribution" could be the title of my last three months. I sent my thing to dozens of directories and listing sites. The pages exist, some even look nice, and the total number of humans who arrived from them is close enough to zero that I stopped writing it down. The only channel that has produced anything real for me so far is embarrassingly unscalable: writing comments and replies in places where builders already talk, and having actual conversations. A few of those turned into people looking at the product properly. No sales yet either, so take this as a fellow zero, not advice. But if conversion is the metric, conversation is the only row in my spreadsheet that is not empty. Did giving the tutorial away in full change who showed up, even if they did not buy?