It's very important to start a new venture with the right set of mindset, passion, execution skills and people.
Those four are like the individual wheels of a car. Any one goes out of sync and the chances of a fatal learning experience rise sharply. Everything that happens before an entrepreneur successfully creates value is honestly the product of decisions that were right or wrong. Which is essentially, and indirectly, what we end up calling learnings.
As much as I've seen, and as a first-generation entrepreneur myself, startups aren't only about the destination. They're equally about the journey and the people in it. A good idea and a good team both matter. Second one is much harder to get right than the first.
These are ten lessons I've come across so far. Each stated plainly first. Then what it actually means in practice. Because a lesson on its own is easy to nod at and hard to act on.
1. Stay connected to the goal, even though the path holds equal importance
The lesson: stay connected to the goal, even though the path holds equal importance.
Both halves matter. Founders usually drop one of them.
Drop the goal and you get a company that's very busy. Every week has work in it. Every quarter has activity. And after two years nobody can say what was actually built. This is the more common failure, because the path is made of urgent things and the goal never is.
Drop the path and you get the opposite. A founder so attached to a specific destination that they can't hear what the market is telling them along the way. The destination they defend is usually the one they imagined before they had customers.
Practical version is a habit rather than a strategy. Write the goal down where you'll actually see it. And once a month ask one question of the last four weeks: did this move us toward that, or was it just work? Most months the honest answer is partly no. And knowing which part is what stops a year disappearing.
2. Be honest and strict with your team. Those are the same thing.
The lesson: be honest as well as strict towards work with your team. This doesn't mean you have to be rude.
People conflate strictness with rudeness because they've mostly seen them together. They're unrelated. Rudeness is about how you speak to someone. Strictness is about what you accept. Kim Scott's Radical Candor framework draws exactly this distinction as its central axis. Caring personally and challenging directly are independent dimensions. Failure mode of an anxious young company isn't "obnoxious aggression" (challenging without care) but "ruinous empathy" (caring without challenging). Sounds gentler. Produces far worse outcomes.
Failure mode in a young company is almost never the rude founder. It's the founder who's so anxious to keep a small team happy that nothing is ever called substandard. Work is quietly redone at night instead of being sent back. Deadlines slip without being named. Founder absorbs it, gets resentful. And eventually the resentment arrives all at once in a conversation nobody saw coming.
What being strict without being rude actually looks like: the standard is stated before the work starts, not after it arrives. Feedback is about the work rather than the person. Given immediately rather than saved up. And it's the same standard for everyone. Including the founder. A rule you break yourself isn't a standard. It's a preference.
Being honest early is a kindness. Being vague for six months and then frustrated isn't.
3. A good idea fails on the wrong people faster than on the wrong market
The lesson: the idea may fail if you choose the wrong set of people, or don't put the effort in the right way.
This is worth being precise about. Because the usual startup story blames the market.
An idea is a hypothesis. Testing it needs execution. Execution is people. With the wrong people the test never runs properly, so you don't even learn whether the idea was any good. You spend the money and end up with no information. Honestly worse than a clean failure, because a clean failure at least tells you something.
The other half, putting effort in the right way, is about direction rather than volume. Effort spent on the part of the business that isn't currently the constraint feels productive and changes nothing. In any given month there's usually one thing holding the company back. Working on anything else is a rest. However tiring it is.
4. For the first twenty hires, never hire someone whose reason is the hike
The lesson: for the first 20 hires, do not (I repeat, do not) onboard people who are shifting just because they're getting a good hike.
These are the most harmful people to the system in the long run. And one way or another every stakeholder eventually realises it.
Reasoning is simple. If money brought them happiness, they'd never have wanted to jump ship and build something of their own. Someone who moved for a 30% raise will move again for the next one. In a young company that departure isn't a vacancy. It takes the context with it, because nothing has been written down yet.
There's a second, quieter cost. A person who joined for the number treats the job as an exchange. This much money for this much work. That's entirely reasonable and completely wrong for the first twenty people, because at that size nobody's job has edges. The work that decides whether the company survives is nearly always the work that wasn't in anyone's description.
Your core team should be putting in the same effort you do. If they aren't, you'll feel left out of your own company. Be very selective about who you start the journey with. Early hires set the standard that everybody after them copies.
None of this means underpaying people. Pay fairly. Just don't let the pay be the reason.
5. Moving people out is a decision, not a cruelty
The lesson: don't feel bad if you have to act logically about bringing people in and out of the ecosystem.
It's a sad reality. But there are many people who do everything in their power to worry about work rather than to do the work. Those people can rot the system, and for a while they can stall the thing you're trying to build.
Distinction is worth learning to spot, because it isn't obvious from the outside. Someone worrying about work is visibly busy. Long updates. Detailed objections. Meetings about the meeting. A clear account of why something couldn't be finished. Someone doing the work is often quieter and finishes things. In a small team the first pattern spreads, because it looks like diligence and it's far more comfortable than delivery.
Two things make this decision less painful when it comes. Make it early. Cost of waiting falls entirely on the people who are delivering, and they notice long before you act. And separate the person from the fit. Most people who are wrong for a five-person company are perfectly good at a hundred-person one. Saying that honestly, and saying it kindly, isn't a contradiction.
6. Experiment at the edges, not at the centre
The lesson: don't be afraid to experiment with something new, or expand your services or product portfolio. It should not go completely off track from the idea.
Genuinely innovating, or building and selling something over the top of what you already do, can be very helpful in the long run. The constraint in the second half of that sentence is the whole lesson.
An experiment adjacent to what you already do reuses your customers, your credibility and your delivery capability. So it's cheap to run and quick to read. An experiment far from what you do reuses nothing. Which means it's a second startup being funded by the first. And the first isn't usually strong enough to fund anything.
Useful test before starting anything new: which of our existing assets does this borrow? If the honest answer is none, it isn't an expansion. It's a distraction with a business plan.
7. Don't build on makeshift tech, and put a date on everything
The lesson: don't waste time with makeshift tech or any other gigs. Always define what you want and when you want it.
Makeshift solutions aren't cheap. They're borrowed. The interest is paid later by whoever has to work around them. And the payment is usually due at exactly the wrong moment. When you finally have volume and the thing that was "fine for now" stops being fine.
Second half, define what and when, is the part that gets skipped. A requirement without a date is a wish. A wish can't slip. So nobody notices it has. The moment something goes off timeline, look for a solution immediately, because more delay only puts the goal further away. A week acknowledged early is a week. A week noticed late is usually a month.
Where the work is repetitive rather than judgemental, automating it is usually cheaper than hiring for it. And it doesn't resign.
8. Hire young talent, but build the induction kit before they arrive
The lesson: hire preferably young talent, and before bringing them into the ecosystem keep your induction and training toolkit ready.
I'm sure almost nobody thinks about this. But it makes life really very easy when training and dealing with a fresh mind.
Here's why the order matters so much. Without a kit, every new person is trained by interruption. They ask, someone stops what they were doing, an answer gets given once and is never written down. The company pays that cost again for every single hire. Answers drift, because ten people explaining something from memory produce ten slightly different versions of it.
The kit doesn't need to be elaborate. What we do and for whom. How a piece of work goes from request to delivered. Where things live. Who decides what. And the five questions every new joiner has asked so far. Takes a couple of days to write. Second hire pays for it.
Fresh graduates may look like a headache today, but over the long run they become the first, second and third level pillars of the company. You don't want to buy a pillar and just place it there as if you were assembling a modular building. Pillars that hold weight are grown in place. Which is exactly what the training kit is for.
9. Keep exploring what the market is doing
The lesson: keep exploring what's happening in the market. There is always room for new learnings, and you should never skip this.
Danger for a founder isn't ignorance. It's being one cycle behind while feeling fully informed. Internal work is absorbing. A company can spend two years becoming excellent at something the market has quietly stopped paying for.
This doesn't require a research function. It requires a small, regular habit. Talk to people who aren't your customers. Look at what your competitors have changed rather than what they say. And pay attention to the tools your own team keeps asking for. That last one is usually the earliest signal you get, and it's free.
The skills that actually matter in the age of AI is a view of where that market is currently moving.
10. Protect the founder, because the company cannot outlast you
The lesson: focus on yourself, family, health, mental health and wealth. Any order is fine. These specifics just need to be ingredients.
Reads like an afterthought in most startup writing. And it's the one lesson with a hard mechanical justification. In a company this small, you're a single point of failure. Founder's capacity is genuinely a business asset. And it's the only one with no redundancy behind it.
Two practical notes. Family and health are the two things that give way silently. Nothing announces that they're being spent. And by the time there's a symptom the debt is large. And personal wealth isn't greed here. A founder with no personal buffer makes decisions from fear. Fear is honestly an expensive way to run a company.
A daily routine that actually sticks is the mechanism most founders skip here, because it's the least urgent thing on any given day and the most compounding one over five years.
Protecting your capacity means handing work over, and most of it should go. But a few functions carry information that doesn't survive being summarised by someone else, and those need a different treatment — you keep the exposure and give away the execution. The five things a founder should never fully delegate is the list, and the reason each one costs you a year later rather than immediately.
And one that sits underneath all ten: networking
Networking plays a good role, and it's worth being deliberate about. Part of that is dropping the bad ghost that makes you say the wrong things when they aren't required. A positive attitude is always appreciated. And in a small market it's remembered.
Reason this belongs with team building rather than sales: the people you meet now are the people you'll hire, be hired by, be introduced by and be vouched for by later. Reputation in a small industry travels far ahead of you and is very hard to correct once it has set. More on this in networking: the one lesson I wish I'd learned earlier.
How these connect
Read separately these are ten observations. Read together they say one thing. In the first two years, the team is the company. The idea can be adjusted. The product can be rebuilt. The market can be re-entered. But all of that is done by whoever is in the room. And who is in the room is the one decision that's genuinely hard to reverse.
That's why the hiring lessons carry the most weight. Why the induction kit matters more than it looks. And why moving the wrong person out early is a kindness to everyone still delivering. It's also why the founder's own health belongs on a list about team building. Person deciding all of the above is part of the team. And the only part with no cover.
Same mechanism runs through brand. A company's brand is just the accumulated evidence of how its people behave. Which is why brand building starts on day one rather than when you have a marketing budget.
FAQs
What matters most when starting a new venture?
Mindset, passion, execution skills and people. They work like the four wheels of a car. Any one of them going out of sync raises the chance of a fatal learning experience. A good business idea and a good team both matter, and almost everything that happens on the way to creating value is the product of decisions that later read as learnings.
Who should a startup avoid hiring early on?
People who are shifting roles purely because they're getting a good hike. For the first twenty hires especially, those joiners tend to be the most harmful to the system in the long run. And one way or another every stakeholder realises it. If money brought them happiness, they'd never have wanted to jump ship and build something of their own.
Should a startup hire freshers or experienced people?
Young talent is worth hiring, provided the induction and training toolkit is ready before anyone is brought into the ecosystem. Almost nobody thinks about that in advance. And it's what makes training a fresh mind easy rather than painful. Fresh graduates can look like a headache today and become the first, second or third level pillars of the company later.
When should a founder move someone out of the team?
When it becomes clear they're doing everything in their power to worry about work rather than to do the work. Acting logically about bringing people in and out of the ecosystem is a sad reality rather than a cruelty. Those people can rot the system, and for a while they can stall what you're trying to build.
How do you avoid losing time in the early stage?
Don't waste it on makeshift tech or other gigs. Define what you want and when you want it. And the moment something goes off timeline, look for a solution immediately, because more delay only puts the goal further away. Where the work is repetitive rather than judgemental, automating it is usually cheaper than hiring for it.
Is being strict with a small team a bad idea?
No. Being strict and being rude are different things. Strictness is about what you accept. Rudeness is about how you speak to people. Common failure in a young company is the opposite of strictness. Nothing is ever called substandard, the founder quietly absorbs it, and the frustration eventually arrives all at once.
Key takeaways
Mindset, passion, execution and people are four wheels. One out of sync is enough.
For the first twenty hires, never hire someone whose reason for joining is the raise. Pay fairly. Just don't let the pay be the reason.
Write the induction and training kit before the first young hire arrives. The second one pays for it.
Move the wrong person out early. Cost of waiting is paid entirely by the people still delivering.
Experiment at the edges of what you already do, so the experiment borrows customers, credibility and capability.
Give every requirement a date. A requirement without one can't slip. So nobody notices that it has.
Founder's health and buffer are business assets with no redundancy behind them.
Always remember: your passion brought you where you are today. Don't let an external factor come in and change what you set out to build. Go for it, and don't stop until it's done as you wanted it. Or as the customers wanted it.
Related reading: brand building: five lessons from working inside SMEs, networking: the one lesson I wish I'd learned earlier, and a daily routine that actually sticks.
