Roast my landing page, by the numbers.
Paste your URL. Eight plain checks score the page, and it's matched against what real tester reports say goes wrong on indie apps. No AI opinion and no invented score.
Free, 2 roasts a day, no signup.
Indie developers test each other's apps here. Their reports are the data.
- No share image, so the link unfurls as bare text
- Price isn't mentioned anywhere on the page
- Clear, short headline
Didn't know what to do first
Create your account to get started
Question for your testers: what did you expect to see right after signing up?
Why this isn't another AI roast
Most roast tools send your page to a language model and print what comes back. That can be useful, and it can also be confidently wrong, because the model has never watched anyone use a product like yours. This one starts from people who have.
Step 1
Real testers review other indie apps
- A finance appPricing
- A habit trackerOnboarding
- A browser gameBugs
- A SaaS dashboardNavigation
Step 2
Every report is tagged and counted
Live counts from real reports, updated as more are read.
Step 3
Your page is matched to the patterns
yourapp.com
Create your account to get started
Didn't know what to do first
A spot worth checking, with a question for your testers.
Where the score comes from
A typical AI roast: Whatever number the model writes down.
This tool: Eight checks in plain code. You can see each one.
What the critique is based on
A typical AI roast: The model's opinion of your page.
This tool: What real tester reports on indie apps say goes wrong most.
Each point it makes
A typical AI roast: Can sound right about any page.
This tool: Has to quote your page word for word, or it's thrown out.
What happens next
A typical AI roast: You get a list.
This tool: Real people can test the same spots, if you want them to.
What real tester reports say goes wrong
Every approved test on IndieAppCircle is a written report from an indie developer who used someone else's app. We read them and tag each one by what went wrong. The roast uses these numbers, narrowed to your platform where there's enough data.
Share of reports that mention each problem. A report can mention several. Shown in aggregate only: no app, tester or quote is ever named, and a group needs at least 10 apps before the roast uses it. Testers stay anonymous, so the faces are illustrations.
The 8 checks, and how to fix each one
Quick, automated and deliberately simple. They read the HTML your server sends, so they can miss things a person would catch. That's what the testers are for.
- A headline a stranger gets in a second
What it looks for. An H1, or failing that the page title, of 14 words or fewer.
Why it matters. Visitors decide whether to stay before they scroll. A long or missing headline means they decide without knowing what this is.
How to fix it. Say what it does and who it's for, in one short line.
- Readable without JavaScript
What it looks for. Real text in the HTML your server sends, before any script runs.
Why it matters. Apps shipped as one JavaScript bundle, common with Lovable, Bolt or a plain React setup, send an empty page. People see it fine. Link previews, many crawlers and AI assistants see nothing.
How to fix it. Prerender or server-render the landing page. The app behind the login can stay client-side.
- Set up for phones
What it looks for. A viewport meta tag.
Why it matters. Without it, phones draw the desktop layout zoomed out, and the first impression is squinting.
How to fix it. One line in the head: <meta name="viewport" content="width=device-width, initial-scale=1">.
- A share image
What it looks for. An og:image tag.
Why it matters. A launch post lives or dies on its link preview. Without an image, your link on X, LinkedIn, Slack or Discord shows as bare text.
How to fix it. Add a 1200 x 630 image and point og:image at it.
- The price, or the word free
What it looks for. A price, a pricing link, or the word free in the page text.
Why it matters. What does it cost is one of the first questions a visitor has. Pricing confusion shows up in real tester reports even on free apps.
How to fix it. State the price, or that there's a free tier, on the landing page itself.
- One clear next step
What it looks for. Action wording such as Get started, Try free, Sign up or Download.
Why it matters. A convinced visitor still has to know what to click. No clear action is where interest quietly ends.
How to fix it. One main button above the fold, labelled with what happens when it's clicked.
- Some proof
What it looks for. Testimonials, reviews, user counts or similar.
Why it matters. A new visitor is looking for a reason to believe you. With no proof at all, the page asks for trust it hasn't earned.
How to fix it. One real quote or one real number is enough to start. Invented ones cost more than they win.
- A person behind it
What it looks for. A contact address, an about page or an email.
Why it matters. When there's no way to tell who runs a product, some visitors assume nobody does. Trust comes up in real tester reports too.
How to fix it. An about link or a contact email in the footer.
How a roast works
Paste your URL
Your landing page, or an App Store, Google Play or Chrome Web Store listing. We fetch it the way a crawler does, without running JavaScript.
Eight checks run
Plain code, pass or fail. The score is the number that pass, nothing else.
Your page meets the tester data
The page is matched against what real testers report about apps on the same platform. When the AI part runs, it may point to up to three spots, each quoting your page and each tied to a real share of reports.
Real testers check those spots
If you want. One click drafts a listing from your page, and the questions from the roast become the test instructions.
Questions about the roast
Is the landing page roast free?
Yes, and there's no account. You get 2 roasts a day. Roasting the same URL again within 7 days returns the saved result and doesn't use one.
Is this an AI roast?
Partly, and the page says which part. The score and the eight checks are plain code. The tester statistics are counts from real reports. An AI model only writes the 5-second read and picks up to three sticking points. Each one has to quote your page word for word and is tied to a real statistic, or it's thrown out. The model never produces a number.
Where does the tester data come from?
From approved test reports on IndieAppCircle, written by indie developers who test each other's apps. Each report is read and tagged by what it's about: onboarding, bugs, pricing and so on. The tool only shows shares across groups of at least 10 apps. No app, tester or quote is ever shown.
How is the score calculated?
One point for each check that passes, out of 8. Nothing else goes into it, and the AI has no say in it.
Why does it say my page isn't readable without JavaScript?
Your server sends a nearly empty page and JavaScript builds the rest in the browser. People see it fine. Link previews, many search crawlers and AI assistants read only what the server sent. Prerendering or server-rendering the landing page fixes it.
Does it work with App Store or Google Play links?
Yes. It reads the store listing like any other page and uses the tester data for that platform.
What do you store?
The result, for 7 days, so a repeat roast is instant. It includes a few short quotes from your page. To enforce the daily limit we keep a one-way hash of your IP address that changes every day, never the address itself.
What happens after the roast?
Nothing, unless you want it to. If you do, the button drafts a listing on IndieAppCircle from your page, with the roast's questions as test instructions, and indie developers try your app and write up what happened. You earn the credits to pay for that by testing other people's apps.
Then let real people try it
The roast shows where to look. Testers show what's actually wrong. List your app on IndieAppCircle for free, test a couple of other apps to earn credits, and indie developers will use yours and write up what happened.
Before you launch
- The pre-launch checklist for indie apps
Week by week, from positioning to launch day.
- Testing before a Product Hunt launch
What to have tested before launch day, and by whom.
- Getting constructive criticism
Hear what's wrong while it's still cheap to fix.
- Verified testimonials
Real tester quotes for the proof section your page may be missing.
- More free tools
Everything here is built on the same tester data.