Shipped my AI tool 3 weeks ago. 0 paying users. Here's what I'm doing wrong.
I'm a backend dev. 10 years of platforms, Elixir, PHP and Go. Last year I went all-in on AI apps.
Three weeks ago I shipped JobHunting. The idea came from watching friends prep for interviews with tools that either throw a generic question bank at them, or worse, tell them to "craft a compelling leadership story" about experience they don't have. So I built something that does the opposite — you put in your actual resume and one job posting, and it shows you where you'd get exposed and what each interviewer is likely to dig into.
Here's the part I'm bad at: it's live, it works, and almost nobody's using it.
What I've figured out about myself over the last 3 weeks: I can ship, but I don't naturally think about distribution. I built first and figured I'd "tell people later." Later is now, and I don't really know how to do the telling part.
So this is me trying. If you've shipped something and hit the same wall, I'd genuinely like to hear how you got past it — not the success-story version, the actual messy middle.
If you happen to be job hunting and want to try it, I'm not going to pretend the signup isn't a goal. But what I actually need more than users right now is one honest answer: when you see the fit report, does it tell you something useful, or does it feel like the same generic feedback every prep tool gives?
Live: https://jobhunting.whataicando.site
Sample report (no signup): https://jobhunting.whataicando.site/demo
I'm the maker. Ask me anything, including why I thought S$9/mo was the right price (still not sure).
Your demo page is the distribution channel - it already sells without a signup wall. Drop it where your exact users already hang out (interview-prep subs, Blind, LinkedIn) and the answer to your real question - 'does it feel generic?' - will come faster than any pricing debate. Let a few hundred strangers run the demo before you touch the $9/mo.
The distribution problem may be real, but I'd test one other thing before trying to send more traffic: whether the subscription model matches the job-hunting job-to-be-done. Interview prep is episodic. Someone can get a genuinely useful report and still not want another monthly subscription.
I'd run a 10-person manual cohort before scaling distribution: pick one narrow role, ask each person for a resume plus an actual job posting, deliver the report, then measure three things: did they open the interview map, can they name one concrete insight that changed how they'll prepare, and would they pay again for another application/report? If most get value but balk at S$9/mo, test one-time report credits or bundles before changing the product. If they can't name a specific useful insight, pricing is a distraction and the report/map still needs work.
I'd also instrument landing -> sample -> upload -> report opened -> map used -> return within 48h. Since you said the map is what makes the differentiation click, I'd make map interaction the activation event rather than signup. Then “0 paying users” becomes a debuggable funnel instead of one number.
You asked whether the fit report reads as useful or as the same generic feedback every prep tool gives. Your landing page already answers that, and the answer is not the one you want.
Your words: "when you see the fit report, does it tell you something useful, or does it feel like the same generic feedback every prep tool gives?"
The headline promises "not a generic question bank", then the proof underneath is strengths, risks, deal-breakers, likely questions and follow-ups — the exact nouns every competitor uses. The claim is differentiation; the evidence is a category description. Worse, the sample report does not contain a report. That page is 504 characters and lists what the sample shows rather than showing it, while saying No signup required. So a visitor cannot judge the one thing you are asking strangers to judge. And S$9 appears nowhere on the site — you are asking whether the price is right, but nobody who lands has seen it.
I read the page as nine sections, mark where a visitor stops believing, and rewrite the copy for those sections. Diagnosis and replacement English copy only — no design, no code, no pricing decision made for you.
The section-by-section map and the rewritten copy come after we start. The three above are free.
No client case study to show — my before-and-after demos are my own samples and I label them that way. What I can point at is this reply: I opened your site and your sample page and named three specific things instead of saying the copy looks clean.
Paste one real fit report into the sample page, redacted, and repost the same question here — if strangers still call it generic after seeing an actual report, the problem is the report and not the page, and that is worth knowing before you touch pricing.
Three fair hits, and the second one stings because it's true — the sample page describes the report instead of being one. I built it as a teaser and didn't notice I'd made the one thing I'm asking people to judge impossible to judge. That's on me, and it's getting fixed this week before I push the post anywhere else.
On the pricing: it's on the landing in the pricing section, but you're right it's buried — a stranger skimming the hero and the sample wouldn't see it. Noted.
On the copy using the same nouns as everyone else — also fair. The harder thing to name is what actually feels different when you use it: you upload your resume and a job posting, and it tells you the specific thing in your background an HR screener will latch onto versus what a future teammate will. That's harder to put in a headline than "strengths and risks," which is probably why I took the lazy route.
I'll handle the copy and the sample myself — appreciate the offer though.
The interesting part is that you already have a much sharper problem than “AI interview prep.” Showing where someone’s actual experience is likely to get challenged feels substantially more useful than generic advice. I’d be curious whether people understand that distinction immediately from the fit report, or only after seeing the demo.
Honestly, naming it that way is something I've been circling without nailing. "Where your specific experience will get challenged" is a sharper frame than "AI interview prep" — I think you're right.
To your question: I don't think people get the distinction from the fit report alone. What makes it click is the interview map — switching from the hiring manager's view to the cross-team partner's and watching the same gap get probed from a different angle. The report says what's weak; the map shows how it gets tested. Right now the page leads with the report and the map is a second click, which might be the wrong order for making the distinction land.
That distinction is pretty compelling. If the map is what makes the problem click, it might be worth seeing whether putting it earlier changes how people understand the product.