What 27 Views Taught Me About Building Without a Community

Of everything I’ve built and sold, the SaaS boilerplates were the ones I was proudest of. Not the Obsidian plugins, not the n8n workflows. SchemaLens, SubSaver, FreelanceFlow, KeyMint, MetricMint. Real backends, real auth, real data models, built over a few days each instead of a single afternoon. In my head, these were “the real work.”

They pulled 27 Gumroad views. Total. Across all five.

The number that actually stuck with me

I’ve already written about the six Obsidian plugins: 204 views combined, 2 sales, about 35 dollars, split in a way that surprised me between the plugins that clear an outside pile (52 views, 2 sales) and the ones that organize your own notes (152 views, zero). I sat with that 152-views-zero-sales number for a while, because it felt like the more damning one. A product nobody clicks isn’t proven bad. A product 152 people click and nobody buys, that’s a real signal.

But 27 is the number I keep coming back to now. Not because it’s a worse number than 152. Because it’s a different kind of number entirely, and it took me longer to admit that difference than it should have.

Two different failures, and I was treating them the same

152 views and zero sales tells you something about the product: wrong pitch, wrong audience, wrong price, something in the listing itself didn’t land. That’s a positioning problem, and positioning problems are at least fixable. You rewrite the description, you retarget the audience, you try again.

27 views across five products tells you almost nothing about the products, because almost nobody saw them to judge. The Obsidian plugins ride on Obsidian’s own plugin directory, a crowd already looking for exactly that kind of tool. The boilerplates had no directory, no marketplace, no community feed, nothing putting them in front of a stranger who might want one. Their zero wasn’t a verdict on the work. It was a verdict on where I’d put the work.

I knew this, intellectually, from the day I wrote about the plugin split. I said as much in that post. What I didn’t do, for weeks afterward, was let it change what I built next. I kept treating “the boilerplates didn’t sell” as evidence about the boilerplates, when the honest read was “the boilerplates were never actually shown to anyone.”

Why the pride made it harder to see

I think the reason I was slower to act on 27 than on 152 is exactly because I was proud of that work. A product built over days instead of hours felt like it deserved a market on the strength of the build alone. It doesn’t. Nobody buys something they never see, no matter how well it’s built underneath. I’d have caught this faster on a product I cared less about.

That’s a specific kind of mistake, and I don’t think it’s rare: the more effort you put into something, the more you want the effort itself to count as distribution. It doesn’t. Distribution is a separate job from building, and no amount of pride in the build substitutes for it.

What I actually did about it

Eventually, not immediately, I pulled the buy buttons off all five boilerplates and pointed them to email and GitHub instead, which I wrote about at the time. That wasn’t me giving up on the products. It was me admitting that a Gumroad link with no traffic behind it was a worse signal than no link at all, and that I’d rather be honest about what those five products actually are right now: real, working software with no live storefront, not failed launches.

Where I’ve put the energy since is the side that already has a discovery layer doing part of the work for me: the Etsy shop, and the plugins that sit in Obsidian’s own directory. Not because the boilerplates were bad. Because 27 views was never a fair test of anything, and I’d rather build the next thing somewhere a stranger can actually find it than build a sixth boilerplate and get the same 27.

I wrote the numbers down honestly in I built six Obsidian plugins. Two sold., and the decision that followed from this exact gap in why I took the SaaS boilerplates off Gumroad. This is the piece in between: the number sat there for weeks before I let it mean what it actually meant, and pride in the build was most of why.