I quit my job on April 5th of this year, so it’s been 8 months now since I’ve been full time trying to build a startup.

Looking back on these 8 months, I realize now that I had no idea what I was doing when I first quit my job. My first ideas made no sense (especially business-wise). They were just interesting product ideas that had literally no grounding in business, revenue generation, or any form of distribution strategy–neither product-led nor with straightforward marketing. I also didn’t have a clear mindset around what I was aiming for, which made my exploration meander aimlessly.

Throughout these 8 months though I’ve been super fortunate to have met a bunch of peers and mentors in the startup space. Ones who are pretty unabashed in their advice and critiques about how I’m fumbling about. The majority of these mentors and peers are ones that I met through EF, so I do wanna give credit where it’s due.

I’ll straight away take a learning that EF drilled a lot: “Strong Opinions Weakly Held”. The way I interpret this is to: quickly form convictions, operate and test these convictions against reality, and be ready to pivot from these opinions. It’s a mistake to delude yourself into ever thinking you’re 100% correct. But it’s also inefficient to operate without conviction.

Now I present a collection of philosophies around the mindset of building startups I’ve started forming strong opinions on. Weakly held, so if you think I’m wrong you must let me know. I hate forming wrong opinions because I didn’t see the full picture.

Quick Note

These philosophies are only applicable if you intend to build a startup business. For the definition of a startup, I’ll just point you to Paul Graham’s definition.

I don’t think there’s a distinction in mindset between Bootstrapping vs. VC. But of course there is a distinction between building a startup meant to scale vs. a traditional business or side hustle.

Don’t Wait For Permission

I see a lot of people who don’t commit to something until “the conditions are right”. For example:

• I’m waiting until I find a cofounder
• I need funding
• Once I get into YC/[insert other accelerator] I’ll be able to [something]
• I can’t quit my job yet
• I don’t have the skills

First I want to address each one specifically:

• Start now!! Cofounders have a tendency to reveal themselves as you gain traction
• No you don’t (except certain scenarios if you’re starting a deep tech startup). 99% of MVPs/POCs are extremely low cost because the point is to get concept across
• If you need the conditions to be correct to start, then you’re not going to succeed. The nature of startups is that one knows for certain that the conditions are correct. If it were certain it would’ve been done already
• Then don’t quit your job yet. Working on startups is gonna be a fuck-ton of work hours, so might as well start getting acclimated to it now lol
• You’ll be forever up-skilling while building a startup. You may not be the best, but you can’t afford to not do important things because you can’t be bothered to learn

Overall these all revolve around the same thing: conditions will never ever be ideal, that’s the game you play with startups. If you need to do something then do what it takes to make the conditions work.

Ambition & Urgency

Something else that’s lifted directly from EF (thanks EF!).

I believe that mindset-wise, ambition and urgency are the only 2 things that matter. For each, there’s an emotional reason as well as a pragmatic reason why these are important mindsets to hold.

Ambition

I’ll add in the caveat that your interests must be aligned with the ambition, otherwise it won’t work out.

Emotionally speaking, it’s tough to work on something that isn’t awesome. As much as I can’t stand the guy (especially on Twitter, which I’ll never call “X” btw), there’s a reason why people will work for Elon’s companies even when the hours are rough (by western standards) and the pay is below industry average. It’s because the goals of his companies are fucking awesome. Across his companies he’s: advancing clean energy infrastructure, building generalized brain-computer interfaces, and attempting to send humans to mars.

Look, I know as much as anyone else that these “North Star” missions may not be the reality of the companies’ operations and/or how they actually make money. But things will always get tough when building a startup. And when those tough spots inevitably hit, it’s much easier to convince yourself to push through when the outcome is huge, and your interests are aligned with wanting to see that ambition play out.

For me personally, when I was working on a few of my micro-SaaS ideas in the beginning and things were starting to get tough it was really hard to convince myself to push through. I’d ask myself “do I really really care about this??”. And the answer was always “not enough to push through this rough patch”, and the momentum would just fizzle out right then and there.

Now my ambition is to change Americans’ habits and routines around their health, because as a nation we’re so unhealthy. I really don’t care if things are tough. This is a tough-to-achieve ambition, but if it’s attained then holy shit we’re cooking. I’m willing to push through to get there.

Now pragmatically speaking, I’ve already eluded to it but it makes hiring (or finding a cofounder) so much easier. Very smart and ambitious people want to work on hard and ambitious things. Shocker I know, but it’s true. This is why Elon’s companies attract so many smart people, and why a lot of them also leave and start their own companies. Like the amount of startups started by former Tesla and SpaceX employees is astounding.

Urgency

Urgency is simpler.

Emotionally and pragmatically speaking, it’s pretty much impossible to land on the exact correct idea and execution on your first try, or second, third, etc.

But if you do your job right, then you’ll learn something with each try and have the potential to iterate towards a more correct idea and execution.

I believe the right state of mind is to be desperately trying to figure out why you’re wrong or what’s correct. The faster you can do this process of having a theory about what might be correct, testing it, evaluating the results, and coming up with a new theory, the faster you can iterate towards something that is more correct.

Even if you’re an industry veteran and already have “good taste”, people are very different from each other. Again, shocker I know, but it’s worth saying because it’s easy to forget. For example I’m personally very into health & wellness. I was the founding engineer at a remote personal training company. And yet many many of my initial hypotheses of my initial customers were completely off base. Even the initial ICP I went after doesn’t make sense now in retrospect.

Having this urgency to iterate also creates a mental separation with your actions and outcome correctness. Your purpose is to discover truth. Not necessarily to be correct every time. It’s easy to hit a low when certain experiments don’t pan out how you expected, but that’s not what you should be urgent about. I find that keeping a clear mindset around urgency on iteration helps keep experimentation momentum high which keeps emotions well-regulated.

I worked on deep-tech for a bit while at EF, specifically foundation models for video-based physics simulation, and even with deep tech you need to hold yourself to a very tight iteration schedule. Here’s my thoughts on iteration speed for the following:

• Any form of software: at LEAST 1 experiment per week, ideally 2 per week. If you’re B2C then stay on the faster end, if you’re B2B stay on the slower end
• Any form of deep-tech software: at LEAST 1 experiment per 2 weeks, ideally 1 per week
• Any form of hardware: at LEAST 1 experiment per month, ideally 1 per 2 weeks
• Any form of deep-tech hardware: these are the true moonshots–likely you get a handful of experiments. Good luck 🫡

Talk to Customers

Unless you’re building a deep tech startup, or you are building a very specific enterprise SaaS where your purchaser is whatever role you most recently had at a similar type of company (e.g. you were recently VP of product at a mid-sized B2C operations-heavy software company and you are now building an enterprise SaaS tool where someone pretty much exactly like you would be the final decision maker on the purchase) then, until you have people actively paying you, your only job is to figure out what your customers want.

This can look differently depending on both what you’re building and maybe your style. But personally since I kinda like building and I build fast, I talk to my customers, prototype something that fits what I learned, put it in front of those same customers, make them actually use it, collect quantitative and qualitative data, and update my assumptions.

The amount of customer conversations you have, and the speed at which you can test, learn, and re-test will be the difference between succeeding and failing.

Distribution Distribution Distribution

The quote has been beaten to death but I’ll repeat it anyways: “First time founders are obsessed with product. Second time founders are obsessed with distribution.” - Justin Kan.

Not that I’m a second time founder but I do feel this very acutely now. I’ll point out some very salient examples.

There was another startup called Ness that did something very similar to what I’m trying to do now. They called it quits in 2023, but that’s besides the point for now. Their cofounder/CEO had previously founded a health media company, essentially a blog and newsletter with >2M readers. Ness started by emailing the >2M people on their subscribers list to kickstart the user base for this new company. Even if someone could somehow magically build a much better product, how do you compete against the ability to immediately interview, test with, and monetize from 1,000 users from day one. That’s so powerful. I wish I had that lol.

Most people would agree that US health insurance is a really shit product. Most have a negative NPS Score. For reference YC suggests that new consumer-facing startups should aim for a minimum of 50. But health insurance has insane distribution. You pretty much automatically enroll in health insurance if you have a job or you are the dependent of someone who has a job. And most Americans have jobs, or have family members that have jobs. So unfortunately, they have insane distribution and still make egregious amounts of money.

Final example. AWS may not be an order of magnitude better than GCP, but they do have an order of magnitude more users. They have a very strong on-ramp from personal developers, startups, and small companies, capturing them very early on. And now the community of people with tribal knowledge about AWS is massive. So if people are choosing a cloud provider, and they just want things to work with help from online forums, and they want those juicy free-tier credits, then you can’t really go wrong with choosing AWS. The actual functionality of the AWS product is not even really all that important in most cases.

I used to not really understand when people asked “what’s your go-to-market strategy?”.

Build something that people want. Also build the way that people will discover that they want what you build.

VC is a Tool

I feel like far too many companies take VC funding when they aren’t exactly really VC-scale startups or don’t actually really need VC funding.

The history of how we got here seems to be:

• There were a lot of successful startup exits in the late 2010’s
• Both financially and clout-wise VC firms did very well
• They raised hella from new LPs
• Now they need to actually deploy those funds
• More startups are more capital-efficient now, so requiring smaller rounds
• VCs need to invest in more startups to properly deploy the capital

And so now we’re here, where the narrative has convinced everyone to take VC funding.

However after a few conversations with a few firms across the spectrums of location, industry focus, and size, my overall sentiment is that VC is a tool. And tools are meant to solve particular problems.

In the case of VC, the problem to be solved is how to blitz-scale a startup such that a proper scale is reached in order to achieve defensibility. Here’s a few examples that fall quite cleanly into this framework:

• Social networks. The defensibility comes from “all your friends already on here”
• Food delivery/ride sharing. The defensibility comes from the gig worker workforce size directly correlating with speed of delivery/pickup time
• Deep tech. The defensibility comes from capital investment required to build the product and eventually it’s not worth it for other companies to build this internally. This worked for NVIDIA through CUDA (though it’s starting to crumble) and didn’t quite seem to work for OpenAI because there wasn’t any lock-in mechanism

Now obviously there’s more examples, and also examples that fit less cleanly into this framework but still work well with the VC model. But overall the idea here is that scale is a large part of what makes the company work well, and more scale generally means more good. So scaling up fast and quickly capturing large parts of the market with proper lock-in means that competitors can’t really play.

I’d be very careful about why a startup should take on VC funding. VCs know what their tool is really good at, so they expect that outcome. If you take on VC funding you need to be prepared to deliver scale, defensibility, and a large part of the market captured without much room for your competitors to play.

A lot of the 2010’s startup success stories happened to be in winner-take-all markets, at least at the time based on whatever lock-in mechanisms the startups used. But not every market is like that. Some markets are so big that it doesn’t make sense. Some markets just simply don’t have that dynamic. And some products fundamentally do not operate on lock-in or becoming drastically better with scale.

Bonus: Don’t Listen to VC Rejection (or Validation)

The only people you should be listening to are your customers. That’s all.

Everything is Cringe Before it’s Cool

My best friend once said that “everything is cringe before it’s cool”. And I think he’s absolutely right about that. It’s cringe to work hard without much result to show for it, that’s why people used to brag in school about how little they studied. Because imagine how cringe it would be if you studied hard but still failed.

You need to abandon that mindset.

The majority of startups are gonna fail even with a fuck ton of effort. And it’s going to feel horrible that you’re putting in so much effort but there’s nothing to show for it for so long. I’m still in that stage, even if I’m getting a steady stream of users, I don’t feel like I’ve hit any inflection points in growth.

Aside from talking to customers, my philosophy now is that the number one job of a founder is psychology management. You need to keep yourself disciplined and focused on ambition & urgency. You need to manage your psychology and not doubt yourself, get into analysis paralysis, and just keep hypothesizing, testing, iterating.

To that end I find that surrounding yourself with the right people is extremely helpful. Find friends who are also founders, ideally actively in the trenches also trying to find PMF. If not that then find friends or date someone who understand the necessity of climbing cringe mountain before anything interesting can be achieved. Content creators like YouTubers and TikTokers actually understand this super well. Doing a TikTok dance when you have 0 followers is hella cringe. But doing a TikTok dance when you have a million followers, now that’s kinda interesting. Even people who would’ve called you cringe from the beginning will start wanting to be associated with you.

That’s all!

Again please–I hate not seeing the full picture and holding wrong opinions. Please please correct me where you think I’m off base.

It’ll be interesting to revisit this next year and see which of my thoughts have changed.