Notes from Future Founder's track
Here are some notes that I've taken down from the future founder's track by YC. Hope they'll be useful to you as they are to me. I also highlighted a few key points.
Why to not not start a startup by Paul Graham #
- Too Young. Age doesn't necessarily correlate to maturity. A simple test would be to see how one reacts to a challenge. Do they immediately dismiss it as a stupid idea or do they probe deeper to understand why.
- Too inexperienced. The best way to get experience running a startup is to start a startup. Getting a normal job for a few years might turn you into someone that cannot move forward without being told.
- Not determined enough. If you sufficiently driven to work on your own projects, you are probably determined enough.
- Not smart enough. If you're smart enough to worry about not being smart enough to start a startup, you're probably are.
- Knows nothing about business. The initial focus is on the product. It must be something that people want, not something you think people want. Don't worry about making money, that's the easy part.
- No cofounder. It's hard to argue against this because a good cofounder is very important. Single founder startups are very tough. The older you grow, the harder it is to find a cofounder within your circle so start early! Young people have nothing to lose and everything to gain.
- No idea. Many startups pivot from their original idea to find product market fit. It also helps if you're good at making things. Find something that's missing in your own life and supply that need no matter how specific to you it seems.
- No room for more startups. Why would there be a limit for something valuable? If what you want to do will solve the need for a group of people better than any solution out there, you have a startup idea.
- Family to support. You shouldn't start a startup if you have a family unless you have the financial runway. If you really want to do it, start a consulting business on the side and turn that into a product business. Best to start when you're young and have no commitments.
- Independently wealthy. Startups are stressful so why do it if you don't need the money? There are many more interesting people doing regular jobs to pay the bills so if you want them as colleagues, you'll have to work at something that pays the bill even if you don't need to.
- Not ready for commitment. A successful startup is going to take at least 3-4 years. If you're not ready for commitment at that scale, don't do it.
- Need for structure. If you need people to tell you what to do, don't start a startup. A startup requires people to know what to do without someone else telling them what to do. Think of it as a soccer team. If you get angry when someone says you're not independent enough, you probably don't need someone to tell you what to do.\
- Fear of uncertainty. If you start a startup, it will probably fail. There is no uncertainty. Hope for the best but expect the worst. It'll still be one hell of a ride, you'll learn a lot and in the best case, you become rich.
- Don't realise what you're avoiding. People who've been working for a year or two knows what they're avoid, they know the regular 9-5 job sucks.
- Parents want you to be a doctor. Parents are generally more conservative for you. They want you to be a doctor not because you save lifes but because it is a prestigeous job that pays well. There's still a correlation between risk and reward.
- A job is the default. Defaults are powerful but without people straying from the norm, we will never make progress. Startups are a rare historic shift in the way wealth is created.
Before the startup by Paul Graham #
- On instinct. If you trust your instincts, you'll probably make a lot of mistakes. It is natural to immediately dismiss counterintuitive ideas but it is not always a good idea. However when it comes to people, it is okay to rely on your instincts. A lot of founders don't do that enough. Work with people you genuinely like and you've known long enough to be sure you can work well with them.
- On expertise. The way to succeed in a startup is to be an expert on your users and the problem that you are solving for them. A lot of young founders go through the motions of startup a startup instead of actually making what people want.
- On gaming the system. We've been trained to find shortcuts to win this game called school since that's what we did growing up. Things that don't actually matter. There's no shortcut, the only way is to make something people want. It really is that simple.
- On commitment. A startup will take time, a few years perhaps. A successful startup will take even more time, maybe even up to a decade. Starting a startup is really hard and it never gets easier. Young successful founders will never get to do the things regular people do but likewise, regular people will never get to do what they do.
- On trying. How can you tell if you're up for this challenge? You don't know. Most of our lives thus far has been pretty smooth sailing. It's nothing like running a startup. Don't assume that your startup is going to be successful because you've managed to ace every (few, artificial, easy) tests you've had in life so far.
- On ideas. The way to get a startup idea is not to try think of startup ideas. The best startup ideas are the ones you wont even realise they're potential startup ideas to begin with. Learn a lot about things that matter, work on problems that interest you and with people you like and respect. This is also how you get cofounders at the same time as the idea. If you get yourself to the forefront of some technology, you will live in the future. Ideas that others think are prescient will be obvious to you because you know what ought to exist.
How to get startup ideas by Jared Friedman #
Common mistakes when generating startup ideas
- Believing you need an amazing idea. Google and facebook were not the first idea. What made them successful were a good enough initial idea combined with great execution.
- Jumping into an idea without evaluating if it's a good one. If your startup is successful, you'll spend years of your life working on it so wouldn't it make sense to spend a couple of weeks to decide what to work on. Think of your initial as a good enough starting point. Good starting points can be tweaked into a good idea. Bad starting points may not necessarily be salvagable i.e. starting over completely.
- Starting with a solution instead of a problem. For example, uber for plumbers. This is a solution, what problem does it solve? Maybe it's hard to find plumbers? If you come up with ideas like this, this is a solution instead of a problem. This is known as a solution in search of a problem.
- Believing that startup ideas are hard to find. There are many problems in the world, the reason you find it hard to find ideas is because you haven't learnt to notice them.
How to evaluate startup ideas
- For each criteria, rate it between 1 and 10 and average your score.
- How big is this idea? How big could it be?. Looking for existing companies that do something similar. It can be small today but it might be very big in the future.
- Founder / market fit? Are the founders expert in what they're doing?
- How sure are you that this is solving an acute problem? Ideally, you have personally experienced this problem.
- Have a new insight? Is there something that you know that others do not know? Maybe something that, on first glance, doesn't sound like a good idea e.g. letting strangers stay in your house for a short period of time but you actually have data to prove that the contradiction is true instead.
Other signs that your ideas might be good
- You're making something you personally want.
- It only recently became possible.
- Successful companies that do something similar.
Filters are bad reasons to reject startup ideas. You need to learn these filters and turn it off. Examples of filters include:
- Hard to get started. Stripe was a very obvious idea yet so many programmers didn't want to deal with learning all the hard stuff that came with credit card payments.
- Boring space. Boring doesn't mean bad idea. In fact, a lot of important things are boring e.g. writing good documentation.
- Seems too ambitious. It is usually the most ambitious ideas that turn into valuable companies. Don't shy away from it simply because it's too ambitious.
- Too many competitors. Better to err on the side on products with competitors. Usually no competitors indicate that there is no demand for this product. Find a market where there are existing competitors but you noticed something that they didn't. If there is no single leader, it means that the existing competition are incumbents.
So how do you actually go about generating startup ideas? Sitting down and thinking about it is not necessarily the best way to come up with ideas. Try and notice them organically instead but in order to notice them, you need to first know how to recognise them. Working at the forefront of some field lets you see startup ideas before others.
- Start with what your team is especially good at. Think of ideas that you have an unfair advantage at execution. Any ideas that are generated this way have founder / market fit.
- Think of things that you wish someone else would build for you.
- What would you be excited to work on for 10 years even if you didn't succeed. This can also lead founders astray because it might be something the founders are very passionate about but might not have a clear path to being a big business.
- Look for things in the world might have changed recently e.g. new technology, new regulation, new problem etc. What are waves you can ride?
- Find companies that have been successful recently and look for new variants of them. Note that these ideas are usually solutions e.g. uber for X. You need to think critically if the idea actually solves a problem.
- Ask people you know for problems they want solved
- Look for industries that seem broken. You need to know a lot about the industry you want to disrupt.
How to get startup ideas by Paul Graham #
The best startup ideas are solutions to problems that you have.
These ideas have 3 things in common:
- it's something the founders want
- it's something they can build
- it's something that few realise are worth doing
Working on a problem that you have is good because it ensures that the problem really exists.
When the startup launches, there must be some users who really need it and will use it, not just people who could see themselves using it in the future.
There are 4 types of ideas:
- something a small number of people want a small amount (avoid this!)
- something a small number of people want a large amount (focus on this!)
- something a large number of people want a small amount
- something a large number of people want a large amount (usually already exists)
A good idea to work on is a solution to a very painful problem that people will use it even if it's a crappy version made by a two-person startup they have never heard of.
Being at the leading edge of a field doesn't mean you have to push it, you can also be a user of it.
Being able to put together an MVP is a substantial advantage over founders who cannot code.
You need to train your mind to be able to notice these ideas e.g. question things that you take for granted / don't normally think about. Note that this also takes time so don't rush it.
Pay particular attention to things that annoy you.
It's weird because coming with startup ideas is about seeing the obvious but it's hard to see the obvious if you haven't seen them.
This is why a frontal attack i.e. sitting down and thinking of ideas may not necessarily be the best one.
Work on hard problems driven mainly by curiosity and always be on the look out for gaps an anonmalies.
Working on multiple domains is a particularly fruitful source of ideas. You might be able to see problems that software can solve.
Don't feel like you have to build things that will become startups, that's premature optimisation. Just build toys that you think are cool.
Worrying that you're late is one of the signs of a good idea. A quick search on google will settle the question.
The success of a startup is usually more about your execution than it is about your competitors.
If you're uncertain, just ask the users.
A crowded market is actually a good sign, because it means both that there's demand and that none of the existing solutions are good enough.
There's no such thing as a market that's obviously big with no competition. To succeed in a big market with competition, you need some kind of secret weapon e.g. Google. If not, find a market that is small but will turn out to be big e.g. Microsoft.
It is imperative that you turn off the schlep (too much work / tedious) filter.
Every sufficiently good ideas will have a lot of schlep.
The next best thing to an unmet need of your own is an unmet need of someone else. Talk to people!
One way to ensure you do a good job solving other people's problems is to make them your own e.g. if you want to solve a problem in the waitering in industry, be a waiter!
A good trick for bypassing the schlep and to some extent the unsexy filter is to ask what you wish someone else would build, so that you could use it, something that you would pay for right now?
How to Talk to Users by Eric Migicovsky #
There's only 2 things you need to do to start your company: 1. build your product and 2. talk to users.
Common mistakes in a user interview:
- Talking about your idea instead of about their lives. You want to extract information about them to improve your marketing / positioning, not to sell them on using your product.
- Talking about hypotheticals (what it could be, what to build, if I build X will you pay for it?) instead of specifics that have already occured in their lives. Ask them questions about their lives and how they arrived to this problem. Learn about their motivations, why they got themselves into that problem in the first place. Don't focus on how your solution solves their problem.
- Talking too much. Listen and take notes instead. Goal is to extract as much information as possible, real data.
Five good questions to ask in every interview:
- "What is the hardest part about doing the thing you are trying to solve?". For an idea like dropbox, "what is the hardest part about working on a school project with school computers?". Notice how you don't even talk about dropbox.
- "Tell me about the last time you encountered this problem?". Extract context about how the user encountered that problem. You want to ensure that your product actually solves a real problem for users.
- "Why was this hard?". What were the specific things that they encountered that made it difficult? e.g. maybe they accidentally submitted the wrong files because they were relying on email to send files back n forth. Customers don't buy what, they buy why.
- "What if anything have you done to solve this problem?". If potential customers are not exploring potential solutions to their problem, it might indicate that this problem is just not painful enough. What tools did they try to use? Did it solve their problem?
- "What don't you love about the solutions you've tried?". What is the differential between your solution and the existing solutions in the market.
3 stages when to talk to users
Got idea? -> Find users with problem. Start with friends, co workers then get intros from them. You can even drop by events etc. Start with yourself. You don't have to talk to 1000s ppl. You just need enough to get an unbiased overview of the problem. Take notes, keep it casual and be mindful of their time. Just react, you don't have to preplan anything, just need to have enough evidence to support that a problem actually exists and that the solution you had in mind (that you haven't shared with anyone) solves it.
Built prototype? -> Find your best first customer. Get numerical / hard answers! How much does this problem cost them / how much did they spend to solve this problem? How frequent is the problem (the more frequent the better)? How large is their budget / how much are they willing to pay to solve it?
Launched? -> Find product market fit. Specifically, iterating towards product market fit. Read this when you're at this stage. If possible, try and get their numbers as well cause maybe you just need to talk to them again. Don't ask users what they want, you need to understand whether or not these features will help make your product more useful. Maybe you can even ask to see who will pay more for this feature. Discard bad data e.g. non specific compliments / fluffs that are very general.
How to pitch your startup by Kevin Hale #
There's a reason why this article How to apply to YC by Paul Graham is shared when you are making a YC application. Read it.
Don't sell investors the idea. They will sell it themselves if they are convinced.
Do investors understand the idea? Are they excited? Do they like the team and want to work with them?
Clarity is most important. Good investors will do everything else.
Marketing and advertising are a tax companies pay for not making something remarkable.
Best companies grow organically by word of mouth.
Legible ideas can be understood by people who know nothing about your business.
Avoid ambiguity e.g. it takes a question to understand what you just said.
Avoid complexity e.g. single dimension, no coupling. Go straight to the point, don't beat around the bush. Simplicity means you are not trying to mix a bunch of things in your description.
Avoid mystery e.g. jargon, "this", "it" etc if it's not defined.
Avoid ignorable e.g. marketing speak, mba speak, buzzwords. Be conversational. Something you can tell to your mom and she gets it!
Make your idea reproducible. Investors need nouns. What are you making? What is the problem? Who is the customer? For example: "Airbnb is the first online marketplace that lets users book rooms with one another instead of hotels." "Dropbox synchronises your files across your / your team's computers."
Idea must be succinct and concise enough to give the investors a very high overview of what you are building. If they are interested, they will ask more questions like how did you do it? is this the right team? how far along are you guys? do you have traction / customers?
Should you use X for Y? e.g. Uber for Y. X must be a household name. It must be a billion dollar idea. It must be successful, not infamous. Does Y want X? Y should be a huge market.
Lead with what, not why or how. The other stuff gets in the way of your clarity.
Investors read a lot of applications. Make each minute count.
Being concise tells the a good investor a lot about the founder.
Be careful when capitalising your one liner description. You might give investors the impression that you're focusing on the buzzwords i.e. a solution in search of a problem.
Nouns are most important. Adjectives are iffy, investors need to ask a lot more questions to know if they will actually give you an unfair advantage.
There are a lot of cool things that you're doing and you want to tell them but investors just don't have time to read everything.
You give yourself the best chance of getting an investment if you help investors and to do that, you need to get them excited and curious so your idea needs to be clear and concise.
When you have to do a demo in front of 1000s of investors and only have 2 minutes to pitch, the goal of the pitch is to get them to want to talk to you after the demo. Try to focus on you instead eg. your presence, how you carry yourself out, how confident are you etc. The slides are there just to make the founders look impressive, use a language that eveyone understands and is clear and feels inevitable. Talk about it in a way that everyone understands.
The yc application is also about finding more about the applicants, how they think so that the yc partners will want to "go out on a dinner date with you".
How to balance between technical jargon pitch vs non technical ones? You have to do both. Different pitch for different groups of users.
The best pitches make investors want to join the company.
If you have to ask whether or not you have product market fit, the answer is probably no. On the other hand, you're so busy and you have so many customers coming in and asking questions and you're too busy to ask questions like "how do I hire", "how do i serve my customers" and you're making money then that's probably product market fit.
How to evaluate startup ideas by Kevin Hale #
Even though this isn't part of the future founder's track, I just thought I'd add it here as well because I found this very useful.
- YC defines startup as a company that is designed to grow very quickly. There's nothing against small businesses, just that investors are generally more interested in those that grow really fast.
- There are good investors and there are bad investors. Bad investors will poke holes in your business, like trying to show how smart they are. Good investors will listen to you, try to envision what you are telling them and pitch your idea back to you as to what they think your business needs to do to become a unicorn.
- A startup needs to have 3 things:
- a problem
- a solution
- an insight
- A problem is like the initial setting. What are the conditions as to why this business will grow quickly.
- A good problem is one with the following:
- popular (i.e. many people have them)
- growing (e.g. growing industry)
- urgent (need to be solved very quickly)
- mandatory (e.g. new regulation)
- You don't necessarily need to have every factor but of course the more the better.
- Your solution is also the experiment that you want to carry out to test your hypothesis (problem) within the aforementioned conditions for it to grow quickly.
- Also don't start with a solution otherwise you're basically a solution in search of a problem.
- Insight is why what you're going to try will end up successful. Insights are your unfair advantages. You need to have at least one and it must be related to growth. The insight is why investors will choose to invest in your company over others.
- There are a few types of insights:
- Founder (a PM at Google is not an advantage. Having a PhD with some patent in the area you want to start your business is. Most founders don't have this and it's okay.)
- Market (a market that grows by at least 20% / yr without having you to do anything i.e. you're currently in a trending market. This is the weakest advantage if it's your only one.)
- Product (is your product 10x better than your competitors. Your users need to be like "oh shit this is so much better than anything else". 2x or 3x is nice but it's not enough to convince investors.)
- Acquisition (customer acquisition paths that cost no money e.g. word of mouth. Paid acquisition is the worst way to grow your company because it's an advantage that quickly dwindles.)
- Monopoly (as your company grows, does it become harder for your competitiors to beat you? Good examples are businesses with network effect or marketplaces or anything with like a chicken vs egg problem.)
- Most importantly, you need to be able to build and test your hypothesis. Nothing matters if you can't build your product.
How to find the right co-founder by Harj Taggar #
- Why have a cofounder?
- Improved productivity e.g. co-founders with complementary skills, bouncing ideas back and forth.
- Moral support i.e. someone to lean on when it gets tough, someone to pull you back to earth when you're delusional. Employee is different from co-founder.
- Many successful companies have multiple co-founders
- You should generally wait to start a company until you have a co-founder.
- If you start alone, you need to have very high conviction of your idea e.g. you know the problem very well and your solution definitely solves it. They can also build an MVP of the solution without a co-founder.
- What to lookout for in a co-founder?
- How they can handle stress i.e. people that you've worked with under stressful conditions / close friends
- Their goals and values must be aligned with you i.e. what their motivations are. This is to avoid conflicting goals in the future. Note that goals change over time as well.
- Complimentary skillsets aren't make or break, better to find someone you trust and enjoy working with.
- Where to find co-founders?
- Friends and colleagues
- The best time to find co-founders is when you're not actively trying to look for co-founders e.g. work on projects with many people. This lets you get a sense of what kind of people you enjoy working with, what kind of skills do you actually have etc.
- Find people with similar interests to you at events / hackathons.
- Asking is a very underrated skill. If the person you want to co-found a company with rejects you, ask him / her for who they want to start a company with and get an introduction!
- How do you test out and know for sure that you both want to start a company together before making the final commitment?
- Sometimes you just have to do it.
- Allocate chunks of time to work together e.g. deadline for product, pitching together. At the end, have an honest conversation to see if both of you enjoyed working together.
- Prior to starting anything, make sure you are in agreement with these 2 things:
- Equity split. Default is 50:50. Not many reasons to deviate from 50:50. Both of you MUST be happy with the equity split and are motivated to work for a long period of time. It's okay to deviate if maybe you're working on it earlier than he is but remember not to optimise for the short term. A great co-founder is worth more than the progress you made by yourself in the last 6 months.
- Who's the CEO. Initially, it doesn't matter since everyone needs to do a lot of grunt work. Just because you're the CEO doesn't mean you can abscone from some of the hard work. That said, the CEO is effectively the face of the company, the final decision maker. This is very important especially when it comes to investors so if your co-founder is the person who will lead the conversation w/ investors then it is better for him to be the CEO.
- Once it's done, formalise it by setting a legal agreement with vesting.
How to work together by Kevin Hale #
- Founders need to optimise for a relationship that will last 10 years.
- Co-founders will also fight:
- Fundraising / Runway
- Customers / Employees
- There are 4 major things to avoid:
- Criticism (bring other issues into play instead of addressing the issue at hand)
- Contempt (intention to insult, making things personal)
- Defensiveness (someone not owning responsibility of the problem)
- Stonewalling (won't engage / talk)
- 4 things that can help to protect us:
- Divide and conquer (to protect against defensiveness)
- Who is responsible for what?
- It's easy to delegate but people don't talk about what happens when things don't go the way you want to i.e. what is also considered success / failure or when the other co-founders should intervene.
- Know yourself
- What is your attachment style?
- What is your co-founder's attachment style?
- Both parties need to know when attachment styles conflict and what each of them need to do.
- Document a process (to protect against criticism)
- Do this when you are sober
- Figure out roles, goals and a process for handling disagreements before emotions enter the fray.
- It's hard to think straight when you are emotional.
- Decision Disagreement Framework
- You just need to agree ahead of time.
- Use non violent communication (to protect against contempt)
- Be honest without putting down others.
, I feel because I'm needing some . Would you be able to ?
- Observation (fact) vs evaluation (opinion) e.g. you are 10 minutes late to this meeting vs you're always late.
- Emotions vs thoughts. If you replace "feel" with "think", it doesn't work e.g. "I feel frustrated" is an emotion vs "I feel / think that you aren't taking this seriously" is a thought.
- Be careful of anger (pissed) or evaluative emotions (e.g. blamed, judged)
- List of evaluative words that are confused with feelings
- Universal need (everyone should have) vs personal need e.g. I need support vs I need support from you.
- Request (invitation that can be rejected) vs demand
- How to deliver constructive feedback in difficult situations
- Divide and conquer (to protect against defensiveness)
- Pay off your emotional debt EVERYDAY. Don't let small problems fester into big problems.
- Practice having level 3 conversations now (lv 1 - data exchange / passing things around, lv 2 - have some emotions and/or personal, lv 3 - relational, happening right now between 2 ppl that is super important)
- Good practice for lv 3 conversations:
- What are our short term goals for the company?
- Are we using the right metrics?
- Are we hitting our goals?
- Is it clear who is responsible for what?
- Do we agree the current division makes the most sense?
- Is our workload distributed in an optimal manner today?
- Do we feel a high levelof dedication and motivation?
- What measures are there in place for providing feedback to one another
- It's very mentally and physically draining to run a startup so if someone fails to keep their side of the bargain e.g. not following the process that was setup when there is a disagreement, just let them go.
- How do you handle very talented co-founders that are assholes? Hard question. Co-founders need to decide on how they want to run the company. If majority agree it is okay for 1 co-founder to act that way then sure but generally YC prefers that this approach of running a company is better than one that is controlled by dictatorship and fear.
- Someone who's a good friend might not be a good co-founder.
- You have to trust them and believe in them 100% if you are going to hand them half the company.
- How do you trial out a relationship? Do a project together. Contracts work too. Can we consistently have good conversations and make each other better without placating (appease) to one another?
- Take turns e.g. do their idea first then do yours.
- How do you strengthen relationships in remote companies? You have to overshare deliberately. You need to be very good at communication and trusting others. What are things you take for granted when working together? These are the things you need to explicitly do in a fully remote team.
- How do you deescalate an existing conflict? Everyone needs to agree that this is not working / affecting productivity. Sometimes people just need time to cool off.
How to Get and Test Startup Ideas by Michael Seibel #
Another one that's not part of the future founder's track but I'm adding it here as well
- You need to have some kind of relationship with the problem e.g. it is a problem you, your friends, your family, your community faces.
- Good to bounce problems (not ideas) off friends and potential solutions.
- Are there some unique insights to the problem that you know and others don't?
- Look at existing solutions that attempted to solve this problem to have a rough idea of what they tried to do and whether or not they are successful.
- When building the MVP, get it out ASAP and don't fall in love with it. You're better off falling in love with your users and problems instead.
- Should also be careful with who you talk to to test your MVP.
- Most important part of the MVP is that it must actually solve a problem.
How to Split Equity Among Co-Founders by Michael Seibel #
- There are some really good notes in the link above so take a look at that!
How to Plan an MVP by Michael Seibel #
- MVP should be something ridiculously simple, something to give to the very first set of users you wanna target, in order to see if you can deliver any value at all to them.
- Launch something bad quickly
- Get ANYONE using your product to see if they get any value out of it.
- Talk to your users and get feedback. Don't wait and spend millions and years trying to build something thinking that feedback is useless beccause product is still a WIP.
- Hold onto your problems and customers tightly, your solutions loosely.
- Pivoting and iterating is not the same. Imagine you have a problem (problem A) and you built a tool (solution A) for it. If solution A doesn't solve problem A, fix it (iterating). Don't look for other problems that solution A might fix. This is SISP.
- MVPs should be very lean i.e. in weeks not months. Sometimes a landing page with a spreadsheet is enough. AirBnB for example didn't have payments when they first launched i.e. you had to pay your host in person.
- The scope of your solution must be extremely limited. The solution is designed specifically to solve that one problem, that's it. Don't try and address everyone and all your potential users' problems.
- What is the core minimum viable features you need to solve the most pressing problem?
- Stripe, AirBnB and Justin.tv all started with something shitty e.g. there's no location view in airbnb, no payments. Justin.tv only had 1 channel, low res video, no games. Stripe's initial payment processing was done in a very startup-py way, few features etc.
- Helping companies to integrate is a good way to find bugs before users found them! Sure it's not scalable but you don't have to worry about this now.
- Very few cases do you need a "heavy" mvp e.g. significant regulations (insurance, banking), hardtech (rockets), biotech (some drug solution), moonshots (boring company). These cannot be done quickly and needs time so an MVP would just be to start with a simple website telling people what you do.
- Nobody ever remembers launches.
- Most important is to get customers so focus on that instead.
- Hard to learn from customers if they don't have MVP to play with.
- Tips on building MVP quickly
- Timebox your spec e.g. if you only gave yourself 3 weeks to build MVP, what can you build in 3 weeks?
- Write your spec. You don't want feature creeps and you also want to track how often your specs are changing.
- Cut your spec. If everything is important, cut the least important one.
- Goal is to get anything out in the world. Momentum to keep anything going is very strong. If you don't have anything in the world, it's easy to delay.
- Don't fall in love with your MVP.
- How do you balance between features between users? Never ask users for features. It's not the user's job to come out with features, that's your job. Your user's job is to give you problems. What they are suggesting to you are ways that they think will solve their problem. Focus on stuff like "What your problem is? How often do you experience it? How intense is it? Are they willing to pay for a solution? Do they know other people who have the same problem?"
- What if I'm stuck in a cycle where MVP keeps changing and I can't launch? Stop doing what you're doing. Nobody says you have to keep changing your MVP. Perhaps you think it's special e.g. perfect then you keep tinkering with it.
- What's the key thing you wanna learn when you get feedback from your users? Does it solve the problem I wanted to solve? You don't even need to ask, you can usually see when your user is interacting with your MVP.
- Those who in the pre mvp stage should just launch something and then figure out timeline / roadmap. Solve the problem right in front of you first, don't count the eggs before they hatch.
- Post MVP, should you work on growth or retention? Both! It's never binary! Startups are multi faceted and so people like to boil them down to 2 dimensions, A or B. You have to juggle multiple things.
- How do you talk to your users if they have a problem they don't want to talk about? e.g. health records. If you are starting a company to solve type 2 diabetes but you don't know anyone who's willing to talk about this problem then how do you even know it's a problem?
- How do you know when you have product market fit? The classic answer that founders really hate is if you're asking, you don't have product market fit. What happens when you have product market fit is that people start using your products so much, you transition from doing anything other than keeping it alive / online. You stop thinking about new features, how to improve conversion through funnels etc. You are literally like "holy shit I don't know how to serve the people are coming to me, I'm pretty sure we are going to die because we have too many users". Almost no one gets product market fit. Someone using your product means you have a user, that is not product market fit.
- What happens if you learn more about the problem and the problem expands as you interact with the users? That is totally fine and is expected. What common mistake founders make is that they think they have to solve the problem for all users. When a user has a set of problems, what's good to figure out if there are multiple people with a similar set of problems. Just ask them! Some founders are not willing to have a small MVP because their vision is big. Customer pitch and investor pitch is very different. You cannot smosh them together or kill one.
How to Launch (Again and Again) by Kat Manalac #
- Launching is not just one moment in time, it is something you continuously do.
- Types of launches
- Friends and family
- Online community
- Request access
- Social media / bloggers
- New feature / product
- Why launch continuously?
- A/B test your short pitch
- See how users respond to your product
- Are you talking to the right users?
- Silent launch
- All you need is a domain name, company name, short description, contact and a call to action.
- Call to action can be subscribe to newsletter, get in touch etc.
- Product hunt also has a very similar feature, can use this to get started if you haven't already have a landing page but a landing page is still most important.
- Friends and family launch
- Once you have an MVP, pitch it to your friends and families.
- Watch them use it and ask for feedback
- Don't stay in this phase for too long because friends and families may not be the users you are targetting.
- Strangers launch
- Approach strangers who are currently experiencing the problem and solve it for them!
- It will help show you if people are willing to download and pay for what you're doing
- Online community (startup sch, product hunt, hackernews, reddit, facebook, Elpha etc.)
- It's worth putting yourself out there!
- Spend time understanding the communities that you want to launch in.
- Write like you talk, don't talk like a marketing robot, no jargon.
- Introduce yourself, what you're building, who you are building for, why you're building it and any interesting insights (esp. hackernews).
- Build elements into the launches to get people to spread the word e.g. help tweet to skip a few places in the line.
- Social media / bloggers
- Getting bloggers who are active in the area you are trying to target can make a huge impact on early growth.
- Avoid paying if you're an early stage startup.
- Pre order (physical)
- Take a look at successful campaigns and how they did it.
- New feature / product
- Stripe and Glossier are good examples
- Stripe is very good at engaging with the community, getting founders to talk to the users, what kind of problems they are solving, blogging about it, pitching to press etc.
- Glossier releases product on a specific cadence and they stick to it.
- Build your own community even if your product is not ready.
- Launching isn't a one moment in time.
- What do you think of launch parties? Try to spend your money as wisely as possible but if you want to have something small for your community, sure why not.
- If you have different ideas and you're trying to different approaches to solving a problem, how do you approach launching it? Just talk to your customers.
- If you are launching on hackernews, what types of companies are better suited? B2B, B2C, Devtools etc? Hackernews is very technical but you'll be quite surprised what are the top voted ideas. Lean towards intellectual curiosity. What are some interesting insights that you learnt that you can share with the hackernews community?
How to prioritize your time by Adora Cheung #
- Time is scarce.
- Time burns money and money keeps a startup alive.
- Low burn rate is still an opportunity cost.
- You need to be very good at identifying and prioritising tasks that are most impactful to your startup's progress.
- Real vs fake progress (remember action faking!)
- Real startup progress is what moves the needle of your startup. At the beginning, that is growth, how to grow your KPIs.
- Primary KPI is almost always revenue or active users. Set weekly goals for this!
- The highest leverage thing you can do is talking to users and building / iterating the product.
- Fake progress include winning awards, conferences, lots of advisors, pitch competitions, press announcements, forbes 30 under 30, networking events, twitter activity, meeting famous ppl, optisising bs metrics.
- Goal is to deliver real value to your users, not delivering points to your ego.
- How to determine if you're prioritising the right tasks? It's quite easy for low value work to creep into your schedule and you need to actively keep this from happening.
- Keep track hour per hour what you were doing. What is the impact before you started doing and how much your KPI grew in a week.
- low value work fulfills our desire to check as many things off as possible and it feels very good but it doesn't help the company.
- Keep a spreadsheet of ideas to move your primary KPI
- Talking to users helps you convert users to customer / revenue, helps you to understand if you're on the right track, helps you to figure out your product roadmap.
- Building product delivers a solution to the users to see if it translate into more customer / revenue.
- Log them into this spreadsheet but do it later because switching between tasks wastes a lot of time.
- How impactful does each task help you to achieve your weekly goals?
- Common mistake technical founders make is to build then talk to users.
- Prioritisation (high, medium, low), Complexity (easy, medium, hard).
- Don't try to do too many tasks at one time.
- How to know if you're prioritising your time well? If you are hitting your weekly goals consistently.
- Write weekly updates and you have to be consistent. Talk about primary KPI (actual vs growth), biggest obstacles to growth, tasks accomplished and impact and what were your big learnings.
- Once in a while, review all your weekly updates. Check for things like if you're learning fast enough, is your prediction of impact well? are you letting a lot of low value work / fake progress creep in? what are consistent blockers to growth? are you completing tasks in a timely manner?
- maker's vs manager's schedule.
- The art of moving fast = decisiveness.
- The beginning of the startup is to move as quickly as possible, to prove that you are building something people want.
- The key is okay with making the wrong choice and learning fast.
- Thoughts of using OKRs at this stage? At an early stage startup, you don't need them. You just need something simple. Here's my weekly goals and the tasks I need to complete.
- If you dont have product market fit and users are churning, what should you focus on? Finding your problem and the solution that you are building for them. Find 10 people that love your product before you try to grow.
- Hardtech startup needs fund to deliver the product for users, what should I do? Focus on getting more customers e.g. presales (demand) / prepaid contracts.
- Don't get ahead of yourself. Do what is most guaranteed to achieve your weekly goal.
- You have to be honest about yourself regarding the current state of the company, your weekly goals and how much time it takes to do something.
- YC advocates not buying ads as a way to get your customers (esp early on) because it is a sign of bad founder market fit because you don't know where to find those first few users.
YC's Essential Startup Advice by Geoff Ralston, Michael Seibel #
- There are some really good notes in the link above so take a look at that!
Do things that don't scale by Paul Graham #
- Recruit users manually. You can't wait for customers to come to you, you have to go to them.
- Be extremely aggressive when it comes to recruiting. Why setup a meeting to do X when you can do it now?
- Start manually then once you verify that the market actually exists, you can switch to something less manual.
- Your startup will be fragile initially. Don't judge your startup with established ones.
- "How big can this company get if the founders did the right things?"
- If you build something to solve your own problem, you just have to find your peers. If not, do an untargetted launch and look out for folks who seem more enthusiastic than others and go after that group of users.
- Don't just acquire users, make them happy for as long as you can.
- This is quite counter intuitive for many founders because they are mostly engineers (not CS trained), they worry it won't scale and they never experienced it themselves before.
- Big companies don't do these because they can't but you can.
- How much attention should you give your customers? Insanely great. Not just very great, it has to be considered pathological.
- What "insanely great" means depends on where along the timeline your startup is at. For a big company, the product is necessarily the dominant one. For an early stage startup, you can give your users an insanely great experience with an early, incomplete, buggy product but with an insane amount of attentiveness.
- Over-engaging with early users is not just a permissible technique for getting growth rolling. For most successful startups it's a necessary part of the feedback loop that makes the product good.
- Perfectionism is often an excuse for procrastination, and in any case your initial model of users is always inaccurate, even if you're one of them.
- Sometimes the right unscalable trick is to focus on a deliberately narrow market.
- There exists a catch-22 if you want to do a hardware startup. You need money for growth but you need growth for money.
- Sometimes you just need to manually assemble those first 100s or 1000s of hardware in order to gain traction. You can tweak the design faster when you're the factory, and you learn things you'd never have known otherwise.
- Sometimes you should take over-engagement to the extreme, pick a single user and act like as if they are the only client. Tweak until you've met their needs perfectly.
- Another consulting-like technique for recruiting initially lukewarm users is to use your software yourselves on their behalf. It teaches you how to feel like to be merchants who might be using your software.
- "You are your software" - When you only have a small number of users, you can sometimes get away with doing by hand things that you plan to automate later. This lets you launch faster, and when you do finally automate yourself out of the loop, you'll know exactly what to build because you'll have muscle memory from doing it yourself. e.g. Stripe's founders manually signed up their merchants behind the scenes even though on the frontend, it seems like it was "instant".
- Don't obsess over launches. Nobody remembers them.
- what you are going to build + the unscalable things that you are going to do to get the company going (usually along the lines of recruit users manually and give them an overwhelmingly good experience)
- The unscalable things you have to do to get started are not merely a necessary evil, but change the company permanently for the better.
- If you have to be aggressive about user acquisition when you're small, you'll probably still be aggressive when you're big.
- If you have to manufacture your own hardware, or use your software on users's behalf, you'll learn things you couldn't have learned otherwise.
- If you have to work hard to delight users when you only have a handful of them, you'll keep doing it when you have a lot.
Startup = Growth by Paul Graham #
I did not read this yet because I do not have any customers to put my product in their hands yet.