The Reality Check before you commit.

Get a reality check on your startup idea.

Market research, founder-market fit, premortem analysis, and what actually happened to more than 10,000 founders before you. A ten-minute check before you build, commit, or spend real money.

Run Reality Check $29

The read

Four questions we answer before you commit or spend real money.

Is the market there, and can you reach it?

The Market Scan reads your market the way an investor would: what customers already pay for and what that says about how badly the problem is felt, whether the ground is moving toward your idea or away from it, and who controls access to your customer. Each read rated favorable, mixed, or hostile, with the evidence shown.

Are you the right founder for this?

No other tool reads the founder. We read you, and your co-founders, against this specific market: what you already know, who you can already reach, and where the gaps are. Grounded in the frameworks investors use to judge teams before there is a product, and honest where fit is thin.

What happened to founders who tried ideas like yours?

Behind this question sits our record of more than 10,000 real, documented founder outcomes, every one tied to a public source. When your idea genuinely matches stories in the record, you see what actually happened to the founders who walked that path, and why. A chatbot can guess. A record can't be guessed.

Is this the kind of idea that gets funded?

The best accelerators have said, in public, what they look for and which mistakes they warn founders about. We built a knowledge base from that published thinking, for Y Combinator and Andreessen Horowitz, and read your idea through it. Every claim cited to the video or essay it came from. Built from 1,696 videos and 535 articles, distilled into 13,858 sourced findings. Not affiliated with Y Combinator or Andreessen Horowitz.

The Advisory Board

Your hardest question, argued from every side.

Some decisions are too expensive to settle with one opinion. Ask a single model and you get a single answer, delivered with the same confidence whether it is right or wrong.

The Advisory Board convenes four seats, one from each frontier lab: Claude Fable 5, GPT-5.6 Sol, Gemini 3.1 Pro, and Grok 4.6. Ask your question, attach anything they should read, including your Reality Check report, and assign each seat one of twelve roles, from the Contrarian to the Behavioral Psychologist to the Customer Advocate. See all roles

Every seat answers independently, without seeing the others. The answers are stripped of their identities and each seat reviews the others' reasoning blind, so nobody defers to a name. A non-voting Chairman reads everything and delivers one synthesis that tells you where the board agreed and exactly where it split, and why. Not one model's opinion: an objective read across all of them.

$5.99 for the standard four-seat board: one seat from each provider. Want a bigger board? $1.50 per additional seat, up to 8 seats in total.

Ask a Board $5.99 → See all roles →

Who Greenlight is for

Different kinds of founders. One avoidable mistake.

Each is about to make a costly, partially irreversible move. Each can avoid it with ten minutes of evidence.

First-Time Founder

You have a full-time job. You earn well. You watch peers quit and build and raise, and you have an idea of your own. Your coworkers and partner think it sounds great. AI tools make building feel cheap. You're one weekend away from drafting your resignation.

"The decision you're about to make is worth six figures. Spend ten minutes on evidence first."

The Serial Founder

You love building products. You've built before. Maybe even more than once. Months of work, a launch, and then the worst answer a market can give: silence. Not a no, not feedback, nothing. Now you have a new idea, and the old plan: build first, ask later. AI makes the next build even faster. Whether anyone wants it is still the question you skip every time.

"You've already paid for one silence. Ten minutes of evidence before you pay for another."

Foreign Founder

You're building from outside the US. Cheaper, faster, with talent you trust. You believe your product fits the US market. Maybe you're already planning the flight. But the US market doesn't behave like your home market, and you haven't asked it anything yet.

"Don't move countries before you have a greenlight from the one you're moving to."

The real reason

Every founder has heard "validate before you build." So why does almost no one do it?

Startups don't die because founders skip validation out of ignorance. They die because by the time validation would matter, the founder may no longer want the answer. Somewhere between the first spark and the first working version, curiosity hardens into commitment - and AI just made it happen in a weekend. We call it the Commitment Trap.

Read the full thesis →

Which one sounds like you?

The traps that look like progress.

Six patterns of validation that aren't validation. Each one feels productive in the moment. None of them are the market.

The AI-Validated Founder

You asked a chatbot if your idea was good. It gave an enthusiastic yes.

"A chatbot will never tell you your idea is bad."

The Inner Circle Prisoner

Your family is supportive. Your friends are encouraging. Or one skeptical voice killed an idea that might have worked.

"Your friends and family are not the market."

The AI-Empowered Builder

AI made building ten times faster. So you ship before you ask. You pivot based on one customer's reaction, then the next.

"Shipping is not asking."

The Product Perfectionist

You've improved the product six times and traction never came. You assume the product is the problem.

"The market didn't reject your product. It never knew it existed."

The Idea Collector

You generate ideas constantly. You analyze them, refine them, and talk about them. You never test one, because the next idea is always more interesting.

"Stop collecting ideas. Start testing them."

The Serial Pivoter

You've pivoted three times this year. Each pivot felt like learning. But every pivot was based on a new unvalidated assumption.

"You're not pivoting. You're guessing twice."

What we believe

We don't follow the playbook. We follow the evidence.

The playbook says build first and find out later. These six beliefs say otherwise, and every one of them shapes how we read your idea.

Cheap building is a trap, not a gift. AI lowered the cost of building. It did not lower the cost of building the wrong thing.

Emotional commitment is the most expensive thing a founder owns. Once you're locked in, you can't hear the market when it says no.

More than 10,000 founders already ran versions of your experiment. Reading their outcomes takes ten minutes. Repeating them takes years.

Evidence, not a verdict. We show you what we found and where it came from. The decision stays yours, which is where it belongs.

The founder is half the answer. The same idea lives or dies depending on who runs it. That is why we read you, not just your market.

Flattery is free everywhere. An honest read is worth paying for. We would rather lose your $29 than tell you what you want to hear.

For Incubators & Accelerators

Incubators & Accelerators
get a structured signal
before the cohort starts.

Every program admits on instinct. Greenlight puts a Reality Check on every applicant: the market, the founder, and the record of ideas like theirs, before the cohort begins. Optionally read through your program's own methodology, kept confidential to you.

If you're running an incubator or accelerator program and want structured signal on your applicants, here's how it works.

See the program offering →

Get in touch

Questions, partnerships, or just curious?

Reach out if you're an incubator, accelerator, or VC firm exploring partnership opportunities, or if you just want to say hello. We'd love to hear from you.

Thank you. The Greenlight AI team will reply to you shortly.