I put up 3 pages for products that don't exist yet. Each has a real price ($19) and a real "Buy" button. Clicking it opens a box that says the product isn't ready and asks for an email.

That's a fake door test. Here are my numbers after 5 days.

42 visitors. 1 click. 0 emails.

It looks like 3 bad ideas.

It isn't. Not yet.

Why I'm not killing anything

My last product ran 8 weeks.

23 signups. 0 paid.

I spent those weeks rewriting the landing page. The real problem was that I never asked anyone to pay early.

So this time I wrote the rules down before launch:

• Judge each page only after 28 days.
• Fewer than 30 visitors by then? It's a traffic problem, not an idea problem. Extend 14 days.
• Still under 30 at day 42? Close it.
• Buy-click rate under 2%: kill. 2–5%: talk to the people who clicked. Over 5%: build it.
• Never more than 3 tests at once.

By those rules, all 3 pages are stuck at step one: almost nobody arrived.

5 visitors can't tell you anything. 1 click out of 5 is "20%", which is noise.

Read the funnel one step at a time

One total number hides everything.

The idea check page has 29 visitors. Only 9 scrolled down to the price.

That's not "people don't want it". That's "the top of the page didn't hold them". Different problem, different fix.

On my resume tool, 6 of 7 signups in one stretch reached checkout. None paid. If I had only watched "signups", I'd never have seen it.

So I track 4 steps per page:

1. Arrived
2. Saw the price
3. Clicked buy
4. Left an email

Each step has its own fix:

• Step 1 broken → it's distribution.
• Step 2 broken → it's the first screen.
• Step 3 broken → it's the offer.
• Step 4 broken → it's the popup copy or trust.

What I'm doing about step 1

The honest answer: more distribution, every day, for 4 weeks.

• One long-tail guide per page on my own site (Google is slow, so start now).
• Posts like this one.
• Answering questions where the buyers already are.

I'll post the day-28 numbers here, even if all 3 die.

The one I'd like you to try

The idea check is the one I care most about.

You type your idea. It looks at real search demand and at who is already making money in that niche. The free check runs on the page:

https://ideapanner.resuhive.com/?ref=hashnode#idea-check

What's the smallest number of visitors you'd trust before killing an idea?

It's also exactly how you lose an evening.

After a week I had three problems:

1. I couldn't remember what I'd already looked at. Every visit I scrolled past the same cards and re-read them.
2. I couldn't tell why a boring tool makes $3k a month. The card says MRR and a one-line description. It doesn't say who pays, why they pay, or where the customers come from.
3. The filters didn't stick, and I couldn't exclude categories. I never want to see mobile apps, but I can only pick categories to include, not hide the ones I don't want.

So I built a small Chrome extension for myself. It's called IdeaPanner (pan as in panning for gold).

What it does

Mark every card. Hover a startup and you get ✕ Pass · ☆ Later · ◎ Dig in · ✎ Note. Passed startups can be hidden on your next visit, so you only see what's new. The same buttons now also sit under the description on a startup's own page.

Explain. Click ? Explain and an AI writes a plain-language breakdown: what's going on in that market, what hurts, how the product makes money, where its customers come from, whether a solo founder could build it, and where you'd start if you copied it. It reads the startup's public TrustMRR profile, not just the card, so the answer is grounded in the actual numbers.

My list and Compare. Your Later / Dig in startups open in one panel with revenue, your note and the explanation side by side. Switch to Compare and they line up in a table: MRR now, pricing, customer channels, verdict, where to start. It also remembers the MRR on the day you marked something, so you can see what grew since.

Hide categories. Pick Mobile Apps (or anything else) and those cards disappear, and stay gone after a refresh.

Saved filters. Set up TrustMRR's filters once, save them under a name, restore them with one click.

A few things I learned building it

• The site is a React app that re-renders a lot. My toolbar kept vanishing after hydration, and some pages use invisible 0×0 links laid over the visible card. The fix was to attach buttons to the visible card instead of the link, and re-attach whenever the page redraws.
• Detail pages ship two layouts (mobile and desktop, one hidden with CSS). My first version happily inserted its buttons into the hidden one. Only pick elements that are actually rendered.
• Free AI without a server is hard. Asking every new user to go get an API key kills the first click. The new version gives 10 free explanations a day through a tiny endpoint, and falls back to "add your own key" instead of a scary network error if that endpoint is down. Bring your own key (any OpenAI-compatible API) and requests go straight from your browser to that API.

Privacy

Marks, notes and filters live in your browser. No account, no analytics. The free trial sends the startup's public info to my server, which forwards it to the AI provider; it also sees a random install ID and your IP, used only for the daily limit. It only runs on trustmrr.com, and it's not affiliated with TrustMRR.

Try it

It's free: IdeaPanner on the Chrome Web Store. The version with Explain, Compare and Hide categories is in store review right now, so depending on when you read this you might still get the simpler first version.

If you browse TrustMRR too, I'd love to know what you filter on and what you wish the cards told you.

This article was written with AI assistance and reviewed and edited by the ResuHive team.