Startup School 2017

Notes from watching Startup School 2017 class videos.

How to build a product (19 Apr 2017)

"Minimum Viable Data". Don't even have to look at it, just log it somewhere. Have historical baselines.

How to get ideas & how to measure (11 Apr 2017)

How to get ideas

Has a sales team, but not have talked to 90%+ of their customers.

Where to get ideas? Look at own problems as a consumer. Look at all the things you complain about.

Pay attention to other people, and try to observe what's broken.

As a second-time cofounder: external factors better. Easier to raise money, recruit. Not any easier to be successful.


Why did Amazon start w/ books? Differentiation: wanted to do something that would be easy for you but difficult for existing companies. B&M stores have limited space to hold inventory, and books allowed Amazon to allow a better selection than was available then. No headaches w/ licensing & distribution like there would have been with movies, etc.

What's going to be possible that wasn't previously possible?

Measure: users that are getting value today: active users, revenue, transactions.

Cohort: users over a period of time.

"Ring of fire" concept. Rewatch video and better note analogy. You can get huge number of users using your product, but without focusing on existing users you'll end up with no users left. Examples: Groupon, Pokemon Go, etc.

Measure retention.

Measure growth as change increase per week. Want the graph FLAT. Decrease is obviously bad, increase may not be manageable and lead to "ring-of-fire" situations.

Be able to iterate quickly, and try to automate as much as possible such that you can iterate quicky. Be able to measure aspects of iterations. E.g. how long does it take for an engineer hitting commit, unit tests running, and a fix being deployed to production?

Metrics are an abstraction about some other more-difficult-measure thing. Focusing too much on metrics means you will optimize for the metric instead of what the metric is measuring.

Startup Mechanics (10 Apr 2017)

Do a Delaware C-Corp; investors are familiar with it, lawyers are familiar with it. People outside the startup may say an LLC is OK, but it will always be headache later.

Incorporate in the US. US investors struggle to invest in non-US corporations.

Y-Combinator company has many template forms & legal agreements.

How & Why to Start a Startup (5 Apr 2017)

Joining a startup that is already established may make more sense — Facebook Like button, Google Maps, etc were created by later-stage employees and still did fine w.r.t. compensation.

Q: How do you decide whether to work at another company to implement a feature, or create your own company? A: What's your competition against distribution and network effects of existing players?

Cofounders should have a shared history, where you know you can work together. You keep going because you don't want to let your friend down. Don't pick cofounders off the street. Best attributes for a cofounder: (1) values, (2) aptitude & potential, (3) specific skills in order of most desirable skills for a cofounder.

More important to have a few users that love you than a lot of users that like you.

User questions and talking to your users: have you paid me? Why, why not? Why did you stop using our product and use another product? Have you told your friends?

How to get first 100 users: (1) e-mail everyone you already know, ask them. If paid product, actually charge. (2) Research people that may use your product and e-mail them. Conversion rate is low, 2–3%. (3) Social media/forums/etc. Must be an ongoing marketing strategy.

Laziest thing you can do: buy Google and Facebook Ads. Is typically not effective, and is expensive.

Law of compound growth: get 2% better very iteration cycle. Attempt to build fastest iterating company possible.

Flexibility of a company decreases as the inverse of the square of the number of employees.

Resist the urge to hire medicore or inappropriate people.

SamatsWiki: StartupSchool (last edited 2017-04-20 00:09:07 by SamatJain)