Lean Startup Machine requires you to abandon all ego going in. Remember to leave all ego behind going out.

By Grace Ng (Director of Product Experience, Lean Startup Machine)

There are so many nuances to the Lean Startup methodology that it’s easy to think you’re doing it, but you’re not.

Many make the mistake, thinking “Of course I talk to people before I build. That makes total sense. I’m lean.” Although, the concept is easy to understand, it requires self-awareness and an objective mind to effectively maintain it in practice.

I especially emphasize the importance of properly structuring interview questions to get reliable data, and avoiding the traps of confirmation bias. I emphasize these as the foundational structure to starting out lean because this is where I tripped up myself.

2 years ago

I was another young entrepreneur, fresh out of college, trying to get my startup idea off the ground. As a first step, I decided to attend a 3-day workshop on Lean Startup methodologies. At the time, I had no idea what Lean Startup was, it was just recommended as a good starting point for entrepreneurs.

What I took away from that first Lean Startup Machine experience was get out, talk to customers early on and test your riskiest assumption. So that was the only education I took into running my own startup. Sounded easy enough.

My vision was to create a more semantic web. The initial problem I saw was in hiring the right people for personality fit. I started off with a people search engine product for friends to tag each other with skills and characteristics, similar to skills.to and the now defunct, Honestly.

After building out the minimum viable product (MVP), in which we only had the main functionality of tagging friends working, I went out and talked to people. I learned that the motivation wasn’t enough for people to continuously go into an app to tag their friends. So I killed that idea and focused on something else that would give more semantic meaning to search results.


  • Didn’t talk to customers early enough
  • Wireframed and designed the product before talking to customers.
  • Spent more time talking to investors instead of talking to customers.

Time wasted

  • 2 months implementing until the pivot in mid November


  • Didn’t raise $1M to fail.
  • Shaved a few months off by learning from customers early enough.

Pivot to semantic image search

Problem: Textual search is limiting, how do you learn more about the world if you have a visual query you cannot put into words?

Solution: Photo Q&A app that would become a crowd-sourced alternative to Google Goggles.

We built out a website MVP in a few hours, with the one functionality of uploading an image and asking a question. The riskiest assumption we wanted to test was, do people even have visual questions to post?

On my first try, I found one person who enthusiastically responded, “Oh my god, yes! I actually snapped a photo of some kind of bird when I was in Florida. I’ve been wanting to find out what it was, it’d be awesome!” Bingo! This guy actually had a photo sitting on his phone that he wanted to learn more about! The learning was, we had to go mobile first.

The second riskiest assumption was, will there be people who are able to answer the photo questions that are posted? I sent the photo out and, immediately, someone signed up to answer the question, “It’s a Muscovy duck, it’s very common in Florida.” Validation, yet again!

I tested a few more photo questions through emerging photo platforms, Pinterest and Opinionaided. Two-thirds of the questions were successfully answered. It was sufficient validation for us to move full steam ahead, so we went back into the building and hacked away.


  • Fell victim to confirmation bias.
  • Believed insignificant data.
  • Didn’t set a minimum success criteria. If I had determined this, finding one yes out of the few interviewed would not have made it a worthy pursuit.
  • Forgot to talk to customers continuously.
  • Got caught up in features.

3 months in

My co-founder and I had a falling out and parted ways. But he was convinced it was a million dollar idea and went back into his room to work on it for another 6 months before he finally launched it in the Apple App store. Everyone flocked to it and it blew up overnight! No one showed up. No one cared. No ratings. No reviews. Nothing.

Convinced it was a problem in design and marketing, I started up again with another team. Same idea, but more beautifully designed! This time, being more familiar with Lean Startup and being more self-aware doing a startup a second time, I started hearing the feedback I had previously tuned out.

“Um…what would I use this for?” The hesitant response came up over and over again. One guy, after thinking a bit realized, ”Oh…yea… I took a picture of a poster the other day and was wondering what it meant…” Another girl perked up, “Oh cool! So I can take a picture of myself to ask about an outfit I’m wearing?”

“Well, yes…you could…” But that’s not what this is for… I started to see. The problem didn’t seem to exist for people. If it did, it wasn’t painful enough for them to actively be seeking a solution. People weren’t pulling at me demanding to use it. My vision was for the app to be filled with thought-provoking content, a visual way to learn about and tag the world.

Reality was, there was no problem. The three problems that did surface consistently were bird watching, skin conditions (another team validated this idea and won LSM a few months later), and fashion. None of which I was interested in focusing in on. I killed the idea and joined Lean Startup Machine. Since joining, we’ve fleshed out our curriculum to enhance the learning process and train entrepreneurs in the nuances of Lean Startup methodologies through real-world application.

It’s hard to change the way you think. It’s taken me complete immersion in Lean Startup Machine, failed startups, wariness of subjective bias, and emotional detachment from my own startup ideas to actually look at ideas objectively and see what won’t work.

Lean Startup Machine requires you to abandon all ego going in. Remember to leave all ego behind going out.

Biggest mistakes YOU need to avoid

Mistake # 1 – Co-founder didn’t attend Lean Startup Machine, resulting in some lack of understanding. If you have a cofounder or startup team, you should all attend Lean Startup Machine together.

Mistake #2 – Minimum viable product (MVP) doesn’t need to be built. Now we’ve broken down MVP into 3 stages to really emphasize MVP is any experiment you run that will increase your learnings about your customer.

Mistake #3 – Stop selling on vision! Start listening.

Mistake #4 – Don’t ask leading questions. The way you phrase your questions determine the level of reliability in the answers you get.

Mistake #5 – Don’t think or pitch on features of the future.

Mistake #6 – Beautiful design will not cure an ugly baby. (Product that no one wants)

Mistake #7 – Don’t fall in love with your solution and romanticize it in your head.

Savings: I shaved time and $$ off of being unproductive building something no one wants.

Based on my insights from talking to customers, I realized early on that, without a significant problem or passion unifying users, there is lack of structure. This leads to poor quality content being generated.

Interestingly enough, that behavior is being manifested on another photo Q&A app.

This post was originally posted at UXceo.

About the guest blogger: Grace Ng is Director of Product Experience at Lean Startup Machine. Previously, she was an interaction designer at Snapette. Grace is a female designer entrepreneur and interested in helping startup founders understand and craft better user experiences. Follow her on Twitter at @uxceo.