TL;DR

I've co-founded two companies that were both acquired by larger corporations, and I've collected the scars to prove it. The hard lessons cluster around a few themes: you don't always need a co-founder (and a bad one is worse than none), always get a lawyer, validate ruthlessly before you build, talk to your customers obsessively, market in ways that don't feel like spam, and remember that investors fund trajectories, not snapshots. Here's the full field guide I wish someone had handed me at the start.

Co-Founders Are Optional. A Bad One Is a Catastrophe.

There's a myth in startup culture that you need a co-founder, that solo founders are somehow incomplete. After two exits, I no longer believe it.

You can build a real company alone. When I needed skills I didn't have, I hired freelancers on Fiverr and got things done without handing away half my equity and half my decisions. A capable contractor you can fire is often more useful than a co-founder you can't.

And here's the part nobody says at the networking events: sometimes co-founders come with bad intentions. The shared dream can hide a divergent agenda, and you don't always find out until the money or the pressure arrives.

So if you do take a partner, vet them like a marriage. I mean that literally — get to know them on both a personal and a professional level before you tie your futures together. The question that matters most isn't "are they smart?" It's "are they a good person when I'm not in the room?" Because how someone behaves behind your back is who they actually are.

"You don't need a lawyer for this, just trust me."

If a person ever says that to you across a contract, hear it for what it is: a flashing red light. Walk away. I've been burned more than once on paperwork, and every single time the lesson was the same — always get a lawyer. Anyone trying to talk you out of one is telling you exactly why you need one.

Validate Before You Build (It's Never Been Easier)

The most expensive mistake in startups is building something nobody wants. The good news is that it's now almost inexcusable, because validation has never been faster.

With AI, you can stand up a minimum viable product in about five minutes — enough to put a real thing in front of real people and watch whether they reach for their wallets. If building your MVP is taking you more than two weeks, something is wrong: you're over-engineering, gold-plating, or hiding from the scary part, which is showing it to strangers.

Before you write a line of code, though, go talk to people. Here's the validation script I swear by:

1. Find 20 to 30 strangers who actually have the problem you want to solve.
2. Understand how they solve it today — what's their current janky workaround?
3. Find out what specifically frustrates them about that.
4. Ask what they've already tried, so you know the graveyard you're walking into.
5. See whether they're actively looking for a solution, or just mildly annoyed.
6. Ask the only question that matters: would they pay for a better alternative?

That last one separates a hobby from a business. People will say they "love" an idea all day long. The truth lives in their willingness to pay.

"Would you use this?" gets you a polite yes. "Would you pay $20 a month for this?" gets you the truth.

Talk to Your Customers Until It's Almost Uncomfortable

If there's one habit that built both my companies, it's this: I never stopped talking to customers.

Offer them $100 for a proper customer interview. It sounds expensive until you realize that one good conversation can save you months of building the wrong thing. You'll be glad you did, every time.

Then instrument everything. I track customer behavior with Amplitude and Google Analytics, because you cannot fix what you cannot see. Amplitude in particular makes the customer journey legible — where they linger, where they rage-quit, where the friction lives. Watch where people drop off, because that drop-off point is literally your next task.

Use feedback forms for the qualitative side, the why behind the numbers. People are busy, so give them a reason — I offer a chance to win an Amazon gift card, and the responses pour in. Then track the common complaints across both the forms and the analytics, and fix them immediately. Don't let a recurring complaint sit; it's a customer telling you exactly where your money is leaking out.

Two more tools earn their keep. Use UserTesting.com to confirm your UI is at least as intuitive as your competitors' — watching a stranger fumble through your app is humbling and educational. And read your competitors' reviews like a detective; their one-star complaints are a free roadmap of what you can do differently and better.

When People Try to Leave, Don't Just Let Them

Churn is where a lot of founders quietly bleed out, and most of them make it too easy.

When a customer goes to cancel their subscription, don't show them a door — show them a fork in the road. Offer to pause the subscription instead of canceling. Offer a discount. A customer on pause is a customer who might come back; a customer who churned cleanly is usually gone for good.

For mobile apps, keep people coming back with push notifications and a little gamification — streaks, progress, small rewards. Make the app a place worth returning to, not a thing they downloaded once and forgot.

Marketing: Do the Things That Don't Scale

Here is my most expensive marketing lesson, paid for in cash: paid ads don't work unless you genuinely know what you're doing, and I have still never found a marketing agency that got it right for me. If you light money on ads before you understand your funnel, you're just funding a bonfire.

So how do you get your first 100 customers? You do things that don't scale.

Reddit, where real people discuss real problems in the right communities. SEO, the slow compounding kind that pays you back for years. Influencers who actually reach your audience. And a hundred small, manual, unscalable gestures that no growth playbook will tell you to do.

The throughline is respect: people hate spam. You have to find innovative ways to reach them, which starts with genuinely understanding who your target audience is and where they already hang out. Go to their watering hole; don't shout at them from across the street.

And encourage sign-ups even before you charge — capturing emails early gives you a direct line to the people most likely to become customers.

A note on free trials: they work for big companies, but for small ones they're a trap. People will sign up with fake credit cards and burn your resources. Lean on free tiers and discounts instead — people love free things and good deals, and a generous free tier can do more marketing than an ad budget.

How to Actually Get Investors to Say Yes

Let's talk about the thing everyone wants and few understand, because I got this wrong before I got it right.

Investors don't invest in points. They invest in lines.

A point is a single impressive moment — "we hit $10K in revenue." A line is a trajectory — "we were at $2K three months ago, $5K two months ago, $10K now." The number is the same in both stories, but the line is worth ten times the point, because investors are buying the future, and a line is the only honest evidence of where the future is heading.

So before you ever pitch, start drawing your line. Pick the metric that best captures momentum — revenue, active users, retention, whatever proves people want this more each week — and show it climbing over time. A clean upward slope across even a few months says more than any deck full of adjectives.

"How big is this today?" is the question founders prepare for. "How fast is it growing, and why won't that stop?" is the question that actually gets you funded.

Here's how to build toward a yes:

Get into an accelerator. This is one of the best on-ramps to investors that exists. An accelerator gives you a deadline, a cohort, mentorship, and — crucially — a demo day where investors are already in the room, primed to write checks. It converts the hardest part of fundraising, getting in front of the right people, into a structured event. Watching YC's startup videos is, hands down, the best free education on how to start a company and how to pitch one; I learned more from those than from most paid programs.

Show traction, even small traction. Your friends and family might be your very first customers — that's fine, that's normal. Reach out to your network for early feedback before you go wide, and turn those first believers into the opening data points of your line.

Be ready to work brutal hours. Getting a company off the ground demands a stretch of long days that no investor will fund you to avoid. The line you show them is built out of those hours.

Get your pricing right. YC has an excellent video on pricing products — watch it before you guess. Underpricing is one of the quietest ways founders sabotage the very trajectory investors are evaluating.

When you can walk into a room with a clear upward line, evidence of customers who'll pay, and the credibility of an accelerator behind you, the dynamic flips. You stop chasing investors and start choosing among them.

Hire Slowly, and Only Hire People Who Make You Better

A team can make or break a company, so treat hiring like the high-stakes decision it is.

Spend real time on it. Ask the right questions — not just "can you do the job?" but the ones that reveal how someone thinks and how they behave under pressure. Hire people smarter than you; a founder surrounded by yes-people builds a small company. And stay away from assholes, full stop. Talent does not buy back the cost of a toxic person poisoning a team.

Build Fast, But Don't Break Things

"Move fast and break things" is a seductive slogan and a terrible operating principle. Move fast while not breaking things — that's the real goal, and it's achievable.

The secret is tests. Write unit tests and automation so you can ship quickly with confidence instead of crossing your fingers. Hire a QE periodically to hunt for the bugs you've gone blind to. And never make a big change without testing it first — A/B test the things that matter, and let data, not your ego, decide.

Use your own product frequently, too. If you don't enjoy using it, neither will your customers. Dogfooding is the cheapest quality assurance there is.

A Few Truths That Outlast Any Tactic

Some lessons don't fit a category. They're just true.

The important things require money — don't cut corners, or you'll pay double later cleaning up what you skimped on. Keep learning relentlessly, because the world changes and antiquated skills are a slow-motion liability. And keep going after partnerships with bigger players even when it feels hopeless; you may send a thousand emails before one person replies, but that one reply can change the trajectory of the whole company.

If I compressed two exits into a single sentence, it would be this: respect your customers, respect your money, respect the people you build with, and let an honest upward line do your talking. The rest is just persistence — sending the next email, fixing the next bug, having the next conversation, long after the excitement has worn off and only the work remains.