021

How to start

March 2025

Founders often ask how to start a startup. And, I also have a strong personal interest in the topic.

There is an interesting paradox between the time spent on the initial phase (~5% of a startup’s life) and its importance or impact (~50%?). Many decisions are irreversible and costly to fix later. Yet founders don’t spend enough time testing their ideas and foundations. And, don’t realize that working on the right project is more important than working harder on a mediocre one. Too much talent wastes years on fake projects going nowhere. Therefore, ensure the idea is worth many years of the founders’ time and some of the investors’ money. The cost of time is real.

Ultimately, for a VC-backed startup, the main goal is to generate an outlier and a generationally large outcome. Start by “building something people want” in a market that is at an inflection point. Below four points I would consider when starting:

(1) Your “why” will help you survive and strive

The reason why a team wants to solve a specific problem can be a real competitive advantage. It will help to survive dark times when faith is lost and to strive by enduring and compounding over the long term. A strong “why” sets the culture, inspires teams, and attracts talent.

Founders have different “whys.” Usually, money, glory, and fun are finite resources. Impact, solving hard problems, and an insanely ambitious project can be infinite. Building a specific and better version of the future is infectious. 

One thought about money. It can give short-term morale, confidence and validation to the team. But if money is your main goal, aim to become a partner at an investment bank, a VP at a BigTech, or own a profitable SMB. Success and profit are more probable.

(2) The team: the insights quality. Be the right team, not the smartest team

Choosing a co-founder is critical, as co-founder issues often cause startup failure. Build a (1) team of insiders that have valuable insights and that (2) can build a stable and enduring marriage. Note that the insider status doesn’t guarantee valuable or high-impact insights.

First, (1) the goal is to ensure that the team has valuable insights. Usually, people acquire them from a prior job or from a strong personal experience. Either way, ideally, the team that has navigated through the idea maze. This should increase the probability of “building something people want”. Therefore, be the right team to solve this problem, not the smartest team in general. A contrasting view is that a founder with a fast and fresh mind will reduce the gap and outperform by iterating, learning, and executing faster and differently. This is true, but I would still prefer to be an insider with clarity at the starting line.

Marc Andreessen explains the concept of having unique insights or secrets:

“Sometimes, you’re in a conversation and you realize — wow, this person doesn’t have a single original thought. They’re just echoing things they’ve heard from others. And then, there are other people you speak with, and wow — they clearly have their own model of how the world works. It’s a different model from yours, and they say things you don’t expect. And that’s when you realize you need to understand their model versus yours. That’s how you actually learn something fundamental — something that sits right underneath the words.”

Lastly, check the quality of your insights and the impact they can have. Not all insights are equal. Not all insights have a big impact.

Second, (2) co-founder relationships are like long-term marriages and not dating relationships. Partner with someone you trust, respect, and have fun working with. Then invest in keeping the marriage strong and alive. It’s a long-term commitment to yourself and each other.

There are some interesting tendencies on this topic. We often spend weeks vetting a senior hire. But quickly fell in love with the first co-founder enthusiastic about our idea and ready to take risks with us. Another one is to partner solely for complementary skills. In my startup, we were complementary in skills but not compatible in personalities and ways of working. Working together on a previous project is a good way to stress-test this. Lastly, an unequal equity split raises questions about whether co-founders are truly peers. This question matters significantly more to founders than to investors. Also, joining a few months later shouldn’t justify less equity, given the decade-long journey ahead. 

(3) The Idea: focus on the “highest ranked” problem

Good ideas usually start from a problem, are generated bottom-up, and are specific (“narrow and deep, rather than broad but shallow” – PG). The problem comes first, the idea and the product come second, and the business comes third. This explains why good ideas often start as a side project or a hobby. In addition, these ideas look irrelevant to the majority, but are obvious to insiders with secrets. Having valuable secrets is a scarce resource today.

There are frequent pitfalls and misconceptions when looking for an idea. Founders often start a startup because they want to start a company, not because they want to solve a problem. Another is to use a top-down approach – like a consultant or banker looking for a gap or trend in an industry – instead of a bottom-up approach, like a traditional entrepreneur focused on a specific user’s problem. 

Ensure to identify the highest-ranked problem. Those are the problems with the highest level of pain and urgency for users. Many founders identify good problems but fail to identify the highest-ranked ones. This leads to building a nice-to-have solution (a “vitamin”) rather than a must-have (the “medicine” or “painkiller”). A nice-to-have solution will block your growth relatively soon.

A common way to find the highest-ranked problems is at a previous job or through a relevant personal experience. In any case, you should talk to a lot of users. Users hold the truth that founders need to uncover. Getting to the truth of what matters the most is a key characteristic of great founders. 

To learn how to talk to users and understand what matters the most for them, (1) launch early. Launching a mediocre product early is better than an improved product later. It helps you observe users adopting your product sooner. Remember, we all felt ashamed of the first version of our product. (2) Then, talk to users like a doctor. Let them tell you their symptoms and level of pain. But don’t let them decide the diagnosis and the medical treatment. Lastly, (3) rank each problem to prioritize the sequence of your product roadmap. In our startup, we made a simple scorecard listing each problem users face. Then, we included a column for their urgency (the level of pain perceived by users) and one for the frequency (the hours users spend on it). Some common pitfalls when talking to users are confirmation bias and building features that founders want. Remember to “love the problem, not the solution.”

It’s always surprising how much only one or two actions or “wow features” can drive the majority of the results in the early days. The Pareto principle seems overly skewed. I always felt that in our startups, we had a 99:1 ratio rather than an 80:20.

(4) The market: find a shift in the market

Choose a market that can support an outlier outcome while building something people want. Some investors believe great founders find and expand market opportunities. For them, founders come first, due to their valuable secrets and a fast iteration and learning cycle. For other investors, instead, the market comes first. For them, it’s vital to operate in a large market that is growing like a river. Its current can pull the growth of the company. This relates well to a Chinese saying: “If you stand at the wind, even a pig can fly.”

I would focus on building a product that customers want. Instead of chasing fancy currents or trends. Currents are hard to predict and can change over the medium or long term. So, start by solving a high-ranked problem in a large market. 

Something about the market that I wish I had prioritized before is having a clear distinction between SAM and TAM. Founders often target a large TAM but may not realize their SAM is too small and not growing. This can impact growth after Series A. Then, keep an eye on how your SAM and adjacent SAMs evolve over time.

To conclude

I find it exciting to embark on an expedition to discover valuable secrets, and the one or two input that matters the most for a problem and a business. It’s maybe due to a sense of curiosity or a pleasure in doing something new and ambitious. Also, it’s kind of fun to outthink others or at least try to do so. In fact, it’s surprising that competitors and investors study the same markets, customers, and data. Yet, they often reach different conclusions and take opposite actions. Interestingly enough, being a larger and better-funded company isn’t the main driver of success. While having clarity on what matters the most and having the strength to remain focused is.

Ultimately, create something new that people want in an interesting market. Start from the highest-ranked problems. Find them based on your secrets, learn how to talk to users, and keep a short iteration cycle. This is fun and accessible to everyone, but somehow, few can get it right, even repeat founders.

—

Was it helpful? Please share it with other founders and comment below

• WhatsApp
• LinkedIn
• X

—