My Brutally Honest AI Co-Founder
Anyone else get caught in the "I'm sure it's a good idea" trap?
I've been there. I get a cool idea and want to dive straight into building it. Who wants to sit around doing market research when you could be making it happen?
Turns out, that's a shortcut to the startup graveyard.
So, I built a tool to force myself to get honest. 🧭 LeanCompass is an AI-powered sanity check for your startup idea. You tell it what you have in mind, maybe add in a few more details as you go. The AI gives you critical feedback, spots your blind spots, and helps you tighten your plan.
It’s like a candid advisor for those moments when you're just pitching yourself.
That's where I stand now, about 3 weeks in. I have a lot more planned, but no use building before some constructive feedback helping me prioritize where to go.
If you give it a try with your startup idea, do let me know your thoughts!
love this
Thank you! Do you have any feedback or advice for me?
Any startup ideas you'd like to get candid feedback on?
I don't think AI can give a truthful answer, to be honest. Validating an idea is not just about artificial intelligence. It involves conducting in-depth interviews, analyzing competitors, TAM SAM SOM, CAGR, and so on.
Right, AI can't build the business for you (see Anthropic's vending machine experiment). I believe that it can give useful advice, particularly for people with little startup experience.
I'm thinking of it like the sort of mentoring an accelerator provides; you still need to put in the work, but they'll give helpful pointers and candid feedback along the way.
It's true, experienced business development people who know all of those acronyms and have their own playbook won't need it.
The idea and promise had me invest time and made my sign in. But the time and effort invested didn't match my expactations at all - super generic and non-actionable recommendations.
I tried asking in my own GPT account, and told it to be brutally honest about weaknesses with my idea, and it was a far better response.
Meta question: How will you monetize - seems like a one-off thing. You could funnel users into other startup tools and earn partnership/affiliate perhaps?
I love the idea - it's really good, but why should I use the service if I can get a better response with "my own AI" seems to be the general question with these AI driven products?
Thank you very much for taking the time and your candid feedback!
In fact I was just working on the prompt yesterday to make it less generic and more actionable. I've made a copy of your canvas and ran the analysis again.
Do you want to take a look? It's in your account, with "(new prompt)" added to the canvas name.
I'm not sure yet where to find the right balance between too kind and too harsh. If it's too kind it's pointless, if it's too harsh people might turn away in frustration. My current thinking is that the recommendations are actually the most valuable thing since they're very actionable (since the update). What's your take? How would you personally want the feedback to be structured and presented?
Thank you very much again for your effort, I really appreciate it!
You are welcome. I like the tweaked one better for sure.
I was thinking: “Now what?” though.
If you had like: “Click here and I’ll send you 10 easy steps you can throw into ChatGPT” or whatever ☺️
Keep up the good work.
Exactly, that's what I need to work on! I've gotten a few hundred visitors with pretty minimal promo effort (this post, a handful of Reddit comments & tweets), which I take as validation that there is interest among founders.
Next is creating more incentives to sign up (atm only 3% of visitors do) and then to deliver ongoing value, such as sending emails like the one you were suggesting.
Thanks a lot for your advice, do let me know if you have any other thoughts! 😊
This is such a good tool and idea, well done!
Just a heads up since I saw you made an account: I just updated the AI prompt to generate more relevant and actionable advice. I made a copy of the canvas in your account and ran it there.
If you have a moment, would you mind taking a look and letting me know what you think of the change? The canvas is identical, just the analysis is different. You'll find it in your canvas list, with "(new prompt)" in the canvas name.
Thank you! Are you working on a startup right now or in the past? How did you go about validating your idea?
Anyone else get caught in the "I'm sure it's a good idea" trap?
that is swo relatable I'm sure we have all been there
Thank you! What idea are you working on at the moment?
If you've had a chance to try out LeanCompass, I'd love to hear your thoughts!
This resonates so much! It’s easy to get swept up in the excitement of a new idea and skip the reality check phase. I love the concept of LeanCompass—having an AI that gives candid feedback could save so many founders from costly missteps.
Curious—how does it handle ideas that are very early-stage or a bit unconventional? Does it give actionable steps, or more of a high-level critique? Either way, it sounds like a smart tool for prioritizing what actually matters before diving in.
Would love to hear how your first few users have responded!
Right now, it just takes whatever the user gives it. If that has significant gaps, the analysis will mainly point out those gaps and explain why they're important to fill.
I'm thinking of building a "suggestion mode" where it will try to fill in gaps proactively, but not sure yet how to best present that so the user can clearly see what's their own and what's a suggestion.
Right now my priority is on getting some feedback from users. Until now, I've only had feedback from a few friends, who are naturally a bit biased in my favor.
Do you want to give it a try or know someone who should?
This is a fantastic tool that solves a huge and very real problem. The "'I'm sure it's a good idea' trap" is something every founder falls into, and having an AI co-founder for a sanity check is a brilliant concept.
My main strategic question for you is about differentiation. The AI business-planning space is becoming quite crowded. I'm curious, what do you see as LeanCompass's unique "point of view" or "special sauce" that will make it the go-to choice over other tools?
For example, is it the "brutally honest" tone? Or a deep focus on a specific niche (like B2B SaaS, or e-commerce)? Defining that sharp angle seems like it could be a powerful way to guide your next steps.
Really impressive work for just 3 weeks in!
Thank you for the encouragement!
It's true, there is a lot going on in the space. Right now I think it's about giving candid yet constructive feedback, including questioning the user's assumptions.
From what I can see, many other tools take user assumptions as gospel and help implement them regardless of how realistic or verified they may be.
Will that be enough? Do users want that? I honestly don't know yet. That's my own assumption to verify.
Do you have any advice how to go about that verification?
😂 How fitting, this is exactly the sort of AI response I'm trying to avoid with LeanCompass!
Haha, you caught me! 😉 Guess that's the risk when you ask a strategist for their playbook.
Hope the 'AI response' was at least a little bit helpful. Good luck with the project!