I went through Y Combinator Startup school, so that you don’t have to (Part: 1/3)
This blog includes following topics:
How to evaluate Startup Ideas
How to talk to customers
How to plan an MVP
How to launch (again and again)
How to Evaluate Startup Ideas Presenter: Kevin Hale When to say that the current startup idea is good enough that I quit my company. It is a myth that a startup needs to have traction. Lot of ideas at YC are funded which are in the no-code and just an idea phase. How to predict if YC investors will like the idea? Evidence that shows your company can grow quickly Great Investors don’t nitpick on the idea, but he/ she will think about what can make this idea setup for success and pitch it back to you. Startup idea as a Hypothesis has three components:
Problem — Initial condition
What’s a good problem?
Problem is: - Popular: 1M+ - Growing: +20%/ year - Urgent : Immediate - Expensive: $B - Mandatory: Law has changed - Frequent: Hourly
Ideally you should have at least one of the above in your problems.
2. Solution — Experiment that you are running to solve the problem
One advice: Don’t start with the solution
Alway find the problem first and solution later.
3. Insights — Explanation behind the solution
What is your company’s unfair advantage — related to growth?
Have at least one. Good if more than one.
What are Unfair advantage types? - Founder: You are 1/10 people in the world, who can do this. - Market: Industry/ Market is already growing > 20% - Product: Product is 10x better than what exists. - Acquisition: Should get famous by word of mouth i.e. free customer acquisition. - Monopoly: As your company grows, it should get difficult for competitors to beat you.
In order to change someone’s behavior — you need all of these things at the same time.
Two type of beliefs investors have: - Threshold belief — Can they even build it? - Miracle belief (More important) — If you build it, can you make a miracle happen with exponential sales?
How to talk to customers? Presenter: Eric Migicovsky It’s a great situation if you are your own user. This gives you the first step in this journey. Good founders maintain direct connection to their users. No middleman! This is useful in order to extract information from customers for quick feedback and update accordingly. There are only two things you need to build a company:
Write the code (Build the product)
Talk to customers
Parents (mom) are not good people to ask how your startup idea will do. How user interview should look like
Talk about their life, not your idea. Goal is to extract information — not to sell your product.
Don’t talk about specific solution you are building, but general context how they got into the problem at the first place
Founder should talk less. Goal is to extract information.
Good questions to ask:
What’s the hardest part of the thing you are doing? (Ask about the biggest problem)
Ask the last time you had this problem.
What have you done to solve this problem? This tells if the person hasn’t done anything to solve this problem — it might not be a burning problem at all.
What don’t you love about the solution that you have recently found? Users in general are not good in knowing the next best features — but they can tell about the current problem.
Talking to users is most useful at:
Got an idea? → Find users with a problem - Goal: Find first users here. - It is best if you build something for yourself — so that you are user yourself. - Take notes - Keep it casual, so that people won’t feel it is a marketing pitch — which customers hate. - Careful with their time. You love talking about your idea, but they might not do that as much.
Built a prototype? → best first customers. - Ask numerical/ factual questions: - How much does this problem cost to customers? - How frequent is this problem? - How large is the budget to solve the problem? - First user is at the center of this venn diagram.
Launched? → find product market fit - Ask users their phone numbers. To get back for details later. - To actually know if customers are willing to pay — ask them to put in their credit card information. - Discard bad data for ex -general compliments/ fluff → they are not specifics which will help you find usable insights.
These questions will not just make you formalize your solution, but also to create a marketing pitch later when you bring your solution to customers — targeting their problems in hand.
How to Plan an MVP? Presenter: Michael Seibel MVP (Minimum Viable Product) — This is the first thing you can give to the very first set of users that you target — to see if you deliver any set of values to them. You should have vision for every one of your customers, but MVP only for a small set. It’s lot harder for customers to talk about your product, until you give them an MVP Goal of pre-launch startup
Launch quickly (MVP) - Even if you are launching something bad. Do it quickly
Get initial customers - This is very important — Lot of startups die, even before their product interacts with any customer.
Talk to customers and get feedback.
Iterate — continuous improvement on solution - Do this until it solves the problem
Lean MVP (most of the cases)
Very fast to build (within weeks, not months) - Start off with maybe just landing website page.
Very limited functionality
Appeal to a small set of users - Only focus on the highest order problem.
Heavy MVP (very few times)
Here MVP can start with a simple website and video demonstration of what you do.
Example of heavy MVP - Significant regulation (insurance, banking) - Hardtech - Biotech - Moonshots
Launch of Startup
Launch event is not special at all.
No one remembers when Apple/ FaceBook launched
So do it fast!
Launching with customers is much more important than press launch
Hacks to build fast MVP
Time box your spec (Put a deadline) - Do only what you can
Write down a spec (Don’t change spec very frequently)
Cut down your specs - Prune out non-important things
Don’t fall in love with your MVP - There are very high chances customers will make you move on your first set of product features you developed.
How to launch (Again and Again) Presenter: Kat Manalac Launch needs to be very fast, because most of the launches are going to fail. So you need to learn fast and iterate faster. Why Launch continuously?
A/B test your short pitch
See how users respond to your product
Launching to different channels helps you determine if you are talking to the right users.
Types of Launches
Silent Launch - Just have a landing page of your website - Includes basic text information of who you are and what you do. Simple!
Friends and Family Launch - Show them your MVP/ Landing page - Look at their reaction to get early feedback - Don’t spend too much time here, as they might not represent your addressable market.
Launch to Strangers - Show your product to strangers, to get a more de-biased opinion about your product. - Will need to do a little bit more work here, since strangers won’t be as easy to survey compared to friends/ family.
Online Community - Time to get more eyes on your product using online communities - Few of the online communities can be: Startup school, Product Hunt, Hacker News, Reddit, Elpha, Facebook groups, Alumni groups etc. - Start from a smaller online community — take feedback — launch to a bigger online community — repeat! - Write like you talk — don’t use marketing jargon — people hate that. Be cool and casual with it. - Include following details on online community: Introduce your company; What are you building (product); Who are you building for?; Why ?; Interesting insights?
Request for access launch - Start a waitlist to create the hype about the product - Also build viral elements — like member referrals
Social Media - Launch to popular blogs that cover your industry/ product — this is powerful! - First find top blog publications (using simple google search) — then reach out to them to share your story. This way you will get targeted customer eyes on your product. - Some of these opportunities are paid services, but being an early stage founder — you can always find a free one.
Pre-order - Once you have some eyes on your product, raise some money with pre-order launch before you even build the product.
New Product/ New Feature - Every time you launch a new product/ feature — get on social media platforms — create the hype. - This can happen every few weeks.
This blog is part 1 / 3 of the total YC startup school course summarization. Please follow this channel for up to date on coming blogs.