I built a "Startup Graveyard" search engine because ChatGPT’s toxic positivity cost me 6 months and $0 MRR.
Hey IndieHackers,
There is a specific kind of depression that hits you when you launch a product after 6 months of grinding, and the Google Analytics real-time user count stays at 0.
I experienced this earlier this year. Before I wrote the first line of code, I asked ChatGPT for feedback. It told me my idea was "innovative" and gave me a 10-step launch plan. It acted like a cheerleader.
I didn't need a cheerleader. I needed a harsh reality check. We ask AI, "How can I build this?" when we should be asking, "Why is this going to fail?"
I wanted an antidote to "Builder's Hopium." So I built InsightLoop.
It’s an anti-validation engine. Here is how it forces you to face reality through 4 brutal steps:
-
The 100-Round VC Interrogation It doesn't just give you a static report. It initiates a VC-style chat terminal. It asks you about your distribution, your CAC, and why anyone would pay for this. You have to type back and defend your logic for up to 100 rounds until it breaks.
-
Competitor Roast & Moat Teardown We love to ignore existing competitors. InsightLoop forces a lateral comparison. It puts your idea side-by-side with industry giants and explicitly tells you how easily an incumbent could clone your one "unique" feature in a weekend.
-
The Graveyard Database (RAG Matching) Almost every "unique" idea has been tried. InsightLoop runs a vector search against a database of failed startups. If a startup died 3 years ago trying your exact model, it pulls up their corpse and shows you their post-mortem so you don't repeat their fatal mistake.
-
The Collapse Simulation Based on the flaws you reveal in the chat and competitor gaps, it generates a timeline of your inevitable collapse (e.g., Month 3: Traffic dries up. Month 6: Incumbent copies your feature. Month 8: Server costs kill you).
It’s painful to use. But last week, it talked me out of a terrible idea that I was ready to spend my weekend on. The relief of not building something stupid is incredible.
The best part? There is zero friction. You don't even need to create an account or log in to use it. And the entire failed startup graveyard database is free to access.
If you are itching to start a new project this weekend, please put it through the grinder first. Save yourself the 6 months of silent depression.
Link: https://insightailoop.com
Let me know if your idea survives the roast.
I like the idea behind this because it challenges founders to question their own assumptions instead of looking for encouragement.
One thing I found interesting after trying it myself was that the output became much more useful when I treated it as a list of hypotheses rather than a prediction.
For me, the value wasn't "your startup has a 12% chance of surviving." It was "here are the assumptions you should actively try to disprove."
That shifted my thinking from defending the idea to collecting evidence, whether through customer conversations, public discussions, or actual product usage. I think that's where a tool like this becomes genuinely valuable.