1,900 views, 4 clicks, 0 emails: what my fake door test actually taught me

I ran a fake door test days ago and got slapped. Here's what the
numbers actually taught me.

The idea: a tool that enforces a hard spending cap on AWS, GCP and
Azure. The three big clouds have no native hard cap. They send you an
alert 8 to 24 hours after a spike, by which point the damage is done.
The only tool that actually acts instead of just alerting starts in the
hundreds per month, built for teams, not solo devs.

So before writing a single line of cloud integration, I built a landing
page and posted it across a few subreddits to collect emails as a proxy
for "would you actually use this".

The cold shower wasn't a shower, it was a slap.

1,900 views on a single subreddit. One sub. Four clicks on the link.
Zero emails.

Let that ratio sink in, because it taught me more than a win would have.

People didn't go look at what the product solved. They stayed in the
comments and debated the pitch instead. And honestly, some of those
comments were sharper than any click would have been, one person nailed
an objection I hadn't even positioned for.

Three things I took from it:

Engagement and conversion are not the same signal, and they can even be
inverted. Commenting is free and makes you look smart. Leaving an email
costs something. The loudest debaters were the least likely to convert.

The channel matters more than the page. Reddit threads are full of
people in critique mode, not in "I have this problem, help me" mode. The
person who just got burned by a surprise bill isn't in the comments,
they're searching for a fix.

Four clicks means nothing on its own. Even a great landing converts at
2-5%. I didn't have a conversion problem, I had a volume and targeting
problem. You can't conclude anything from four data points.

The question I'm still chewing on, and where I'd love this community's
take: what would actually stop you from handing cloud auto-suspension to
a tool? Is it giving up your keys, the false positive fear that it kills
a legitimate traffic spike, or something else entirely?

(The project itself, if you're curious: https://arc-guard-five.vercel.app/. Feedback welcome, but the learning above is the real reason I posted.)