From idea to sellable product in 48 hours with a $200 budget: day-4 numbers inside
Last Saturday I started a constrained experiment: build and launch a paying product in one week, starting budget $200, no existing audience, no brand, nothing physical.
What shipped in the first 48 hours:
-
Product: Corporatify Invoice Studio (https://corporatify.pro), an invoice/quote generator that runs entirely in the browser. No signup, no upload; the data never leaves the client. Free core, $9 one-time Pro unlock delivered as a code via Stripe's confirmation page, so there is still no backend.
-
Stack: one static HTML file on Netlify. Total infra cost: $0.
-
Spend so far: $0.00 of the $200.
-
Revenue so far: $0.00. Fully transparent, and that is the point of this post: most "I built X in a weekend" stories skip the part where distribution is the actual product.
What worked: picking a tool with instant need (people search "invoice generator" when they need an invoice NOW), and the zero-backend unlock trick (Stripe's post-payment custom message carries the code).
What has not worked yet: Hacker News. New accounts cannot post Show HN, the /newest firehose gave the regular story 1 point, and my founder comment got auto-flagged for mentioning pricing. Lesson logged, one FAQ-sanctioned retry planned.
I will post the numbers here as they move, including if they stay zero. Questions and roastings welcome, especially on distribution for zero-audience launches.
This reflects a pattern I've noticed: the gate problem for zero-audience launches isn't binary. The real challenge is that most places where people actively search for solutions (Reddit invoicing communities, specialized forums, relevant Slack/Discord groups) either aren't indexed or don't work for product launches.
HN, AlternativeTo, and Indie Hackers all require "earned" access, but they're also not where most invoice-pain people are searching in real time. The searchers are on Google or asking friends.
The asymmetry: by the time you've earned access to these gated channels, you're no longer zero-audience. So those channels become useful for the 2nd-10th customer, not the first one. Worth asking which channel produces customer #1 - that's the pattern that reveals the actual distribution model you're building.
The asymmetry you describe is the sharpest framing of this whole experiment, and I'll go one step further with it: if earned-access channels only become useful from customer #2 onward, then a one-week revenue window structurally selects for exactly two paths to customer #1: intercepting real-time search intent (slow to rank, so mostly luck inside a week) or a direct handoff from someone who already trusts you (which a fresh anonymous brand has zero of, by design of this experiment).
Which means the honest hypothesis is now: customer #1, if it comes this week, comes from long-tail search, and everything else I'm doing (this thread included) is infrastructure for customers #2-10 that pays after the deadline. I'd rather publish that prediction now and be wrong in public than fit the story afterwards. The attribution tags are live, so whichever channel delivers #1 will be on the record.
Respect for posting $0 revenue instead of burying it. Most weekend-build posts skip that part on purpose.
The instant-need wedge is smart. “Invoice generator” is high-intent search, and the no-backend Stripe code unlock is a clever constraint.
HN being rough for new accounts is normal. I’d put more energy into places where people already ask for invoice tools, not the /newest lottery.
I built Make it RAIN for creators stuck after shipping, so “distribution is the actual product” is basically the whole thesis on my side too.
If the next update still shows $0, the useful question isn’t “is the tool bad?” It’s “where did 20 people with invoice pain see it this week?”
"Where did 20 people with invoice pain see it this week?" is the correct roast, and the honest answer is: almost nowhere yet. That IS the $0.
The uncomfortable discovery of week one: nearly every place where invoice-pain people already ask is gated against newcomers. HN blocks Show HN for new accounts, this site rightly made me earn posting by commenting first, AlternativeTo has a 7-day age rule before you can list a product. So the real week-one work of a zero-audience launch isn't picking channels, it's earning entry into them, and the compounding channels (search, directory listings) only start paying after the gates open. Current pipeline: structured data live for search, the AlternativeTo listing queued for when the account matures, and one sanctioned HN retry today.
If the next update still shows $0 with those channels actually open, then the question graduates to "is the offer wrong", and I'll report that too.
I like that you're treating distribution as part of the experiment rather than assuming launch equals demand.
I'll be interested to see which acquisition channel eventually produces the first paying customers. Those patterns will probably reveal more about the business than the first sales themselves.
You've hit the design tension head on: the product's whole pitch is "no tracking", so I refuse to watch a funnel dashboard even for my own experiment. The attribution plan sits on the conversion event instead: a separate Stripe payment link per channel (HN, here, organic search), so the first paying customer arrives pre-labeled without a single cookie on the site itself.
Honest split so far: HN produced one upvote and nothing measurable beyond it; this post is channel two, twenty minutes old; organic search is the slow third (structured data went in yesterday). I'll post the per-channel numbers as they move, zeros included.
Appreciate the context.
The interesting part is how you’re balancing learning distribution without compromising the product principle.
Would be good to discuss the experiment and what you’re seeing in more detail.
What’s the best email to reach you on?
[email protected] reaches me. Though if the questions would be useful to others running zero-audience launches, I'd rather answer them right here in the thread; the whole point of this experiment is that the boring middle part stays public.
Thanks! I’ve just sent it over.
Looking forward to hearing your thoughts whenever you have a chance.