The Driest Startup Course from Harvard: 90% of Startups Die at the First Step

📌 Source: https://www.youtube.com/watch?v=q8d9uuO1Cf4

⚡ Core Insight First: The number one reason startups fail is never a bad product — it’s not solving a real and valuable enough problem. This is the core takeaway from a public workshop at the Harvard Innovation Lab. It provides a complete, actionable framework for finding needs and positioning.
All content comes from real startup project critiques on-site — no vague theories. Use it as a checklist before you launch.
It exposes the common self-deception pitfalls of most entrepreneurs. Run through it before starting a project, and you’ll save at least six months of wasted effort.

This is the content of a startup workshop led by mentor Michael Skok at the Harvard i-lab (Innovation Lab). Unlike many startup courses that only point out problems without offering solutions, this class provides a complete, actionable framework from finding needs to positioning. It’s all practical summaries from live project critiques — no fluff.

Many people think startups depend on a flash of insight, funding, or riding a trend. That’s wrong. 90% of startups go wrong from the very first step — they fail to identify the real problem, and they set the wrong direction. We’ve compiled the core takeaways from this session, translating all technical terms into plain language so even a beginner can use them immediately.

🚫 Wrong from Step One: Empty Ideas Are Worthless

Most entrepreneurs get the direction wrong on day one. The core mistakes are two-fold: not identifying the target user, and not finding the real problem.

Michael Skok points out that the number one reason startups fail is never a bad team or a poorly built product — it’s failing to solve a valuable enough problem that meets real user needs. An idea detached from a specific problem is meaningless. Ideas themselves are worthless.

The first rule of identifying your target user: never treat everyone as your target user. Assuming everyone is your customer is a default path to failure. Even global giants do precise segmentation.

What if your user and paying customer are not the same group? For example, the non-profit project Connect-Ed: the users are children in remote areas of Kazakhstan with no devices or digital literacy, while the payers are donor organizations and funders. The conclusion is clear: both matter, but you must start with the users. Only if the users truly find value in the product will the payers be willing to pay.

After finding the right users, the next step is to lock in the Minimum Viable Segment (MVS). In plain English: find a small group of users with highly consistent needs, so you can sell to them repeatedly without modifying your product. This concept pairs well with the familiar Minimum Viable Product (MVP) — an MVP is the minimal functional version built early on to quickly test a need, and MVS is the best companion for testing that need.

There was a team on-site, Compra Facil, building a credit comparison platform. They didn’t even have a clear target user. Michael directly pointed out that this is a fatal startup problem — if you can’t even articulate who you serve, how can you build the right product?

Final reminder: Don’t assume a problem is a real pain point in your head. You can validate a need without spending a dime — go out, interview users, and ask directly. Don’t rush to pitch to users; first, ask them questions.

🔍 Screening Real Problems: Four Filters to Eliminate Fake Needs

If you can articulate the problem clearly, you’re already halfway to solving it. This “Four U Screening Framework” helps you quickly pick real problems from a pile of fake needs.

Michael’s Four U framework consists of four criteria for screening quality startup problems. Only when all four are met is it worth pursuing:

• Unworkable — The problem you’re solving cannot be solved by any current solution on the market.

• Unavoidable — No matter how society evolves, this problem will always exist. Unavoidable needs inevitably create unavoidable startup opportunities. For example, the COVID-19 pandemic directly spawned entirely new industries for masks and testing; the spread of mobile internet directly created the need for mobile banking. These are the best examples.

• Urgent — Your users have limited time, money, and energy. You are always competing with their other needs for priority. Only by hitting their highest current priority can you capture their attention. How do you find their real priority? Simple: when you meet a potential user, don’t introduce your project first. Ask them: “What is the first problem you urgently need to solve right now?” Major market trends themselves create urgency. For instance, the current AI wave naturally brings a bunch of urgent new needs — entrepreneurs can directly ride that wave.

• Underserved — This need is not fully met yet; there is a clear gap in the market. For example, Selene Health, which focuses on menopausal women’s health management. Half of the world’s women experience menopause, but only 2% of medical research funding goes into this area. Relevant products are extremely scarce — that’s a huge market opportunity.

Let’s distinguish two types of business models here: B2B (charging businesses) and B2C (charging individual consumers). Most B2B problems are “pain-type” problems that meet the Four U criteria. Most B2C problems are “desire-type” potential needs. Only by digging into the fundamental needs of the masses can a large platform grow. For example, Facebook, WhatsApp, and Tencent all fundamentally satisfy the human need for connection. The Harvard Study of Adult Development, running for over 80 years, has concluded that loneliness is the number one cause of unhappiness worldwide. The need for connection is always a fundamental necessity.

⚔️ Breaking Through: Startups Shouldn’t Compete with Big Companies on “Faster, Better, Cheaper”

Many startups like to position themselves as “faster, better, and cheaper” than competitors. This is the most dangerous choice. Large companies have far more resources and can crush you effortlessly. Startups need a “3D Breakthrough” to survive.

A 3D Breakthrough means simultaneously meeting three conditions: Disruptive, Discontinuous, Defensible:

• Disruptive: Your approach is completely different from the old ways of the industry.

• Discontinuous: You’re not making small optimizations on an existing track; you’re creating an entirely new one.

• Defensible: You have a moat that competitors cannot easily take away your users.

For example, Google disrupted Microsoft’s software-selling model by giving away its office suite for free and monetizing through ads. Airbnb disrupted the travel industry without building new hotels, by sharing spare rooms. AWS pioneered cloud computing — a classic discontinuous, disruptive innovation. Without cloud computing, today’s mobile internet and post-pandemic remote work wouldn’t exist.

Defensibility is what we often call a moat. Common sources of moats include:
First, network effects: the more users a product has, the more value each user gets, which in turn attracts more users, creating a positive flywheel. For example, WeChat and Instagram have extremely deep moats — even if you build a better new social product, no one will use it without existing users.
Second, high switching costs: users have to invest significant time and effort to switch products. For instance, changing your phone number means changing your carrier. Most users won’t bother. High switching costs are a great moat.
Third, intellectual property, patents, exclusive partnership agreements, and data accumulation. The positive flywheel for data products is even more obvious: more users → more data → better product → more users. The moat deepens over time.

For platform-type products that face the classic “chicken and egg” problem, there’s a solution here: you don’t need to satisfy all needs yourself upfront. Build an open, easy-to-use platform and let third-party developers create features for different scenarios. For example, when the iPad first launched, it was criticized as a “big iPhone” and didn’t sell well. Steve Jobs didn’t build all the must-have apps himself. By opening the platform, third-party developers created apps for various scenarios. Now pilots use it for navigation, surgeries use it to replace traditional keyboards — it became a necessity for different users. Same with the Apple Watch: the first generation was called useless. After opening the ecosystem, it grew core necessities like health monitoring, and now it’s an indispensable product for many.

One final reminder: most products are not complete solutions; they have external dependencies. For example, no matter how good Tesla is, if there aren’t enough charging stations, many users won’t buy. Ignore these dependencies, and you’ll find your product can’t be implemented.

⚖️ Calculate: Is Your Product Worth the Switch for Users?

Users are naturally lazy and reluctant to switch from a product they’re already used to. Only if the benefit your product offers is large enough will users make the switch. The key is to calculate the Gain/Pain Ratio.

The Gain/Pain Ratio is simple: it’s all the benefits the user gets from your product divided by all the pain (cost, hassle, risk) they have to endure to switch to it. The higher this ratio, the more attractive your product.

Most entrepreneurs only focus on the benefits they provide, completely ignoring the hidden pains of switching. For example, switching suppliers in B2B requires changing processes, training, and worrying that your startup might go bankrupt tomorrow. These are hidden costs many never calculate.

What’s the threshold for this ratio? In B2B, a common requirement is at least 10x the benefit to make a company willing to switch from an existing solution. The threshold for consumers is lower, but there’s no fixed standard — you have to ask your target users.

A simple and effective technique for crafting a value proposition: you must clearly describe the extreme difference between before and after using your product. Turn your product from a “nice-to-have vitamin” into a “must-have painkiller.” The bigger the difference, the stronger the appeal. For example, Selene Health for menopause management has a very clear contrast: before, women are plagued by hot flashes and night sweats, unable to work or live normally, even causing complications; after, they return to normal life. The difference is instantly obvious.

How do you uncover the real reasons users won’t buy? When interviewing, don’t just ask “Do you like my product?” Ask repeatedly: “Why wouldn’t you buy my product?” That’s how you uncover the hidden resistance beneath the surface. Many startups die because the Gain/Pain Ratio is too low, users can’t be bothered to switch, and the founder never knows why.

✅ Final Judgment: Great Startups Are a Match Between Founder and Market

After learning the framework, it all comes down to the founder. The core prerequisite for a great startup is that the founder is uniquely matched to the problem they are solving.

Many startups start as “potential aspiration products” — things users want but are nice-to-have icing on the cake. The goal of a startup is to turn that “nice-to-have” into a “must-have,” from a non-essential to a necessity. What is a necessity? It’s something users can’t solve their core problem without. For example, food, water, and health are must-haves. Winter coats are a necessity; the added fashion value is just icing.

How do you make something a necessity? The core is Founder-Market Fit. In plain English: your background, knowledge, and resources uniquely suit the problem you’re solving. Why are you the one doing this, and not someone else? If you can answer that clearly, your odds of success go up.

For example, Taste of Kenya, a project making affordable high-quality coffee locally. The founder could sign a 10-year exclusive deal with local coffee farmers — that’s an advantage no one else has. Even in consumer goods, which are considered hard to defend, you can build a moat. Another example: the women’s health bicycle seat project. The founder has a patent, can accumulate scarce data on women’s cycling health, and has long-term partnerships with bike-sharing companies. That’s natural fit.

Startup success isn’t about a flash of insight, luck, or burning cash on traffic. It’s about finding the right problem step by step and positioning yourself correctly. 90% of startup failures happen at the first step — not clearly articulating the problem, not finding the real need. Before you start a project, go through this framework. It will filter out the vast majority of fake needs and save you at least six months of wasted effort.

💡 Core Insights

• The number one reason startups fail is not solving a valuable enough problem.

• Ideas are worthless; ideas detached from specific problems are just daydreams.

• Treating everyone as your customer is a default path to failure.

• If you can articulate the problem clearly, you’re already halfway to solving it.

• Don’t be a “nice-to-have vitamin” for users; be a “must-have painkiller.”

• Startups shouldn’t position themselves as “faster, better, cheaper” — big companies can crush you instantly.

• Users are naturally lazy about switching. Without a big enough benefit, no one will change their habits for you.

• The core of a great startup is that the founder is uniquely matched to the problem they are solving.