Books In Minutes Uncategorized

Book Recap: The Lean Startup Chapter 3 – Learn

The effort that is not absolutely necessary for learning what customers want can be eliminated.

Eric Ries
The Lean Startup

There are two kinds of learning. One is as an excuse for failure of execution. The other helps make genuine progress in the core metrics of a startup with empirical data, which Ries named it validated learning. Here’s what happened at IMVU that Ries shared in the book.

Ries and the team at IMVU came up with the idea to bring their 3D avatar tech into the instant messaging market. They came up with a brilliant strategy to avoid competition with the already crowded market. After six months of hard work, they finally launch the product. However, no one wants to use it. Now here comes the fun part.

You might already figure out what ingredient they are missing – yes, talking to users. That is exactly what the team did after countless desperation. But not until yet another long round of desperation did they realize what the customers really want.

Ries slowly learned after years that many of the efforts in the early days were wasted. Because those outcomes could have been earned in other forms, just easier. He figured out that they could have conducted an experiment, just offering customers the “opportunity” to try out the proposed features, and more importantly, measuring their behavior. They could have known if it worth the effort to really build the thing. Nowadays, I believe this fairly popular idea is called A/B testing.


How Do I Build This Uncategorized

How Do I Build This? A Podcast Player

Why do I want a new podcast player?

I have started listening to many podcasts every day since last summer, and I feel like Google Podcasts is missing some things that I really needed. Things that would have save me quite a lot of time every day and things that I think should be built into a podcast player. Let me explain.

Moving latest added episodes to the top of the play queue

This is a bad habit of mine. I have a bunch of episodes sit in my play queue, and I just can’t help keep adding more day by day. Among all these interesting episodes, I will finish some of them from some particular shows prior to others first every day, and then choose from the rest based on what I am interested in at the moment. Here’s how it looks like.

I add a fresh new episode of My First Million to the play queue, and scroll all the way down to the bottom to find it, and then drag it all the way up back to the top of the queue. And I do it again for This week in startups. And I do it again for Tech Crunch. And I do it again for 科技島讀. And I do it again for some interesting pods that I learn while listening to another podcast. And I keep doing these many times a day, and each of them takes 30 seconds just like this

Fave a great episode

You keep nodding your head while listening to this awesome episode, and you just want to save it somewhere even though you’re probably not going to listen to it again. But you just can’t find the heart shape icon anywhere!

All those valuable stuffs mentioned in the pod

What was that book the guest mentioned?
What is that guy’s twitter handle?
What was that episode the host just referenced?
Well said! But I can’t remember the exact words…

Yes, the host said it will be put into the show notes, but it’s not there. And not every thing you find valuable is to the producing team. You just hope someone could note that for you…

How will I do this “How do I build this” stuff?

I am going to roll out weekly updates (hopefully) on the stuff I am building. I will build one thing at a time, and it will be related to software at least to some extent. And each stuff will be split into 3+ series containing the following parts

  1. The problem
  2. Design
  3. Build
  4. Launch
  5. Iterate

The problem – This is the hardest part, as a true problem likely will not just come to you, you have to dig it out. Anything I build should be a solution to solve a problem, just like how I lay out the problems that I have for a podcast player in the previous section. The problem should always comes first, and then figure out a solution to solve it. It should not be done in the opposite way, where you make something, and then define what problems it could solve, as Kevin Hale referred to as “Solution In Search of a Problem” in How to Evaluate Startup Ideas.

Design – Too often we skip the design part, and head straight for building the thing. Because it feels so good to just do it. And then we almost always bump into some trivial but cumbersome-to-fix flaws. However, it does not mean we need a thorough 500-page plan. I will present the solution to the problem in this part, and the right fit of technology stack and tools for it.

Build – I will then build it as a minimum viable product, which means to focus on just solving the problem.

Launch – Release it on to the internet. Might be a website, or on to those app stores. But other than that, more importantly, I should take actions to spread the word out. This is my weak spot, and I would be learning by doing it.

Iterate – The above work should not take months and years. It should be short sprints, and I will make improvements on user feedback.

Okay, so the problems I have for the podcast player I’m using right now is there already. Leave your email to receive the next update.

Why do I want to build a podcast player my self?

Because it’s fun. It’s always fun to see what is behind the scenes, and it is my strength to demystify how things work.

Also because I am a Android user, I don’t have access to Apple Podcast. And I want to keep my Spotify account to just music, and back at the time I started listen to podcasts, Spotify did not support changing playback speed. And I don’t really want to try out all those podcast player out there.

Above all, I am on my journey seeking for startup ideas. I have stumbled over the last eight months, and I figured out that I should make it a long run. I am using this series to build audiences, and consistency.

If you have not subscribed yet, leave your email to receive the next update.

Podcast Recap Uncategorized

Podcast Recap: My First Million #157 – Instagram Food Drops Making $200k a Week, Chrome Extensions That are Crushing It & Open Salaries

I have been listening to many pods since last summer, and I decided to write down some recap for those episodes that I think are worth listening to. Let me start with the latest My First Million episode. Read on to learn more about Shaan and Sam’s findings and brainstorms on food, plugin business, and open salaries.

Food Porn

Shaan’s right hand man Ben found three amazing Instagram food brands, and these brands work like this:

  1. The brand announces the flavors of the food for the week on social networks and its website weekly – a.k.a. the food drop
  2. Customers make order on the website
  3. The food brand ships the food in a few days

It is the internet that makes this kind of business possible. These food brands do not own a single storefront, every order is made on the website, and the food is shipped to the customer. They do not keep any production in stock, they make the food after orders come in. Before the internet, a small business can only serve the people around it physically, and it has to open every day to fulfill the constant needs of the same group of people. Now, these small businesses can live a different life. Here’s what makes the different.

Internet is the new storefront. Along with logistics, a small business can reach out to a broader pool of customers. These food brands ship nation wide, and even world wide. And avoid the distribution cost to the retailers. Social networks are the best marketing tools. Along with a little bit of creativity, drop a new flavor that you have never seen before every week and constantly create new hypes. And finally, only take pre-orders online to precisely control the cost.

The three food drop brands:
My Cookie Dealer –
1-900-Ice-Cream –
Allie’s Banana Bread

Sam’s formula for food virality

  • Make a side ingredient the main thing
  • Normal foods with different color
  • Same thing with different size
  • Combine two things that are quite different
  • Make normally crappy food fancy
  • Sub out food allergy

Sam’s idea – find the next Cronut

Have a restaurant at a high traffic area, and launch every three months to sell new, does-not-exist stuff like rainbow cheese cake. See what stuff has the cheapest cost per click, and trademark that thing just like how Cronut did, and license the trademark to make money.

Shaan on opportunities

  • Investing or private equity to scoop up Instagram food drop brands
  • Find out the best of a food category and do food drop for that brand yourself on Instagram
  • Cold chain 3PL that fulfills the booming needs

Sam’s idea – Eat This Not That on Instagram

Bring the old publication on to the internet. Create easy-to-share two column content that has two similar foods with nutrition facts on each side.

Plugin Businesses

Sam talks about a Google Sheet add-on he recently found, because he’s looking for a tool to manage he’s money.

Tiller Money

Most people use Excel to manage their money. Tiller Money is a Google Sheet add-on, which cost $100/y, that pulls your bank account into Google Sheet automatically.


A Google Sheet that imports data from marketing platforms like Google Analytics and Facebook Insights into Google Sheet. Sam guessed it makes 8 figures a year in revenue.

Open Salaries

Salary transparency has long been a controversial topic. It’s good for the employees, but not good for the employers. There’s a secret Google Sheet that has anonymized Google employees’ salaries on it. And Buffer has been totally transparent on all its employees’ salaries.

Database Businesses

Collects salaries among different levels at the big tech companies. And hires former HRs from those big tech companies as freelancers to help users negotiate salary with the companies for a flat fee.


Owned by MorningStar. Collects privately held company data from VCs and put them behind pay wall.

The Org

Collects company’s org chart and put them behind pay wall.

Pay and get sales number a Shopify store is doing. One of its ways is to order from the store and see the order number.

This episode is fun. Food drop business is something I have not heard of in Taiwan yet, I think there will be one soon. Plugin businesses and database businesses would be something worth doubling down. If you have not listen to this episode yet, you should.

Google Podcast
Apple Podcast