Engineers should QA too

I can’t agree more with this post ( by Sriram Krishnan.

If the goal of a product team is to build a high-quality product, then the quality assurance responsibility shouldn’t fall on the shoulders of one or two people. Everyone on the team should be playing and testing the end product.

The reasons for having the engineer be responsible for the quality of the product they build is rooted in psychology. In social psychology there’s a phenomenon called the diffusion of responsibility effect (, where people are less likely to take action when they know others are present.

By having QA, it makes it easier for engineers to take shortcuts on implementation so that they can pass it on to QA rather than to build it right the first time.

By keeping the QA role with the engineer, the engineer knows when his or her job is done.  There is a fast feedback loop for the engineer to learn when his or her decisions works or doesn’t.  The engineer is incentivized to write correct and maintainable code so that he/she doesn’t spend more and more time on QA and less and less time on the writing actual code.

Google Glass – Post #1 – It’s a Brave New World, with Few Roads and Plenty of Potholes

Got my Google Glass yesterday.  I was very curious to try it out because it seemed like a potentially world-changing hardware device.

I’m fascinated by how it can retrieve and store contextual relevant information faster than what I can do with my phone or computer.  The promise of a display that is with me all the time, means more real estate for taking notes and displaying information.  It all worked out great in my mind, but the initial experience has left me wanting more.

Somehow, I’d expected Glass to seamlessly integrate into my daily routine, just like my first iPhone and first computer.  But sadly, it will have to overcome some major design and usability challenges before that happens.

It all started out very encouragingly.  When I first went to pick up Glass at Chelsea Market, I was greeted by a nice gentleman at front-desk and was quickly handed off to a Jill, a Glass specialist.  She was friendly and did a wonderful job walking me through how to wear and use Glass.

The first problem came before I even turned on Glass.  You see, Jill had to make adjustments to the frame to make it fit me just the right way.  She explained that the HUD isn’t meant to be placed right in front of my right eye, but above it, so that I can maintain eye contact. While it makes sense, I think this might feel a little counterintuitive to anyone who tries on Glass for the first time (I’d thought it was suppose to be worn in front of my eyes).  It made me wonder how Glass can proliferate if everyone who buys it has to go through a special fitting.

Once she made the adjustments and we turned on Glass it did leave quite an impression.  If you’re not impressed when something the height and width of a small piece of tape transforms into an 8 feet screen right in front of  your eyes, then you are probably a magician.  It was something very magical and memorable, and at that instant I felt like the door of a new world had been opened.

We walked through some of the gesture onboarding – swiping forward, swiping backward, swiping up, swiping down, tapping.  It’s a subset of motions that smartphone users are used to.  The second problem surfaced when I started playing with some of the basic apps and started confusing which gestures I should use.  Unlike smartphones where the gestures are operated on the screen itself, in Glass the gestures and received from the side of frame and translated to an operation on the HUD.  The separation of input and output, made those gestures very unintuitive.  Two days later, I still find myself trying 2-3 different gestures before finding the right one.  It’s not as intuitive as my first smartphone experience.

I got enough of it though to advance to the next step in my Glass training.  It turns out Glass isn’t very useful without internet, so the very first thing we did was to setup an internet connection. The third problem came up at this point. I was surprised to find out Glass can’t link to the internet directly via cell networks.  Instead, it has to be tethered to a phone or linked to a wifi-network.  While the linking doesn’t take long, a device that is suppose to be omnipresent and retrieve information at a moment’s notice shouldn’t have to rely on other devices/networks for internet.

After I brought Glass home to play with it some more, the fourth problem with the Glass came up.  After about a full 1 hour of interaction (at Chelsea and at home), Glass was ready for some R&R (rest and recharge).  I was surprised at how quickly the battery drained.  I’m used to Apple products that easily go 1-2 days without charging. This was probably the most disappointing find in my first day of Glass.

Even after all of these complaints, I still think Glass can be a device that has some world changing applications. For example, doctors at Beth Israel Deaconess are using Glass to treat patients in a novel way.  I just don’t see it production ready to be worn by millions of people anytime soon.  For now, it is still very much in an infant stage, with many explorers trailblazing in this whole new world.


Major Challenges that Glass Should Address

  1. Battery Life: Current battery life is way too short.  Battery life needs to be extended 10x to last for a full day.
  2. Controllability: Swipe and tap input isn’t fully intuitive.  Voice command often has a lag likely due to internet translation.  In additional, voice seems to only work at the home screen level, even though it’d be more natural to have it work across all levels (after all in real life, we can talk and change our mind at any point).  New input interface likely need to be explored to enable more intuitive control of the interface (maybe in the form of a touchpad watch, eye tracking for scrolling, or finger motion sensing).
  3. User Interface:  The Glass UI is a very simple card based design.  Unfortunately such a design doesn’t communicate depth well and can frustrating to browse through when looking for a particular card in a big stack.  I often found myself lost/disoriented.  There are at least three layers of depth, a) home screen, b) applications list, c) application internal interface.  However, there are no indications which layer of the hierarchy the user is currently on.  This is a problem because depending on the layer the same gesture command can have very different meaning.  New presentation formats and signifiers for user’s place in the OS should be explored so that users can stay oriented at all times.
  4. Latency: With much of its functionality residing in the cloud, latency needs to be significantly reduce to give users a faster input and retrieval experience.
  5. Internet Connectivity:  Glass should be able to connect to internet directly without connecting to wifi or tethering to a phone.  The dependency on other devices creates an unnecessary hurdle and nuisance for users who expect to stay connected without interruption or external dependencies.
  6. Fitting: The fitting process is tolerable, even if the frame form might not be suitable for every form of daily use.  If 1-5 can be addressed so that Glass’ true potential can be unlocked, then I think people would tolerate a fitting.


What successful people do in the morning

I read this post ( about what successful people do in the morning while waiting for my flight back to China. It very much resonated with me.

Last year was the most challenging year of my life. The third year of being a founder was the hardest. I often found myself completely drained by the end of the day. And by weekend, I was ready for nothing but simple rest and recovery.

But of course life has other plans. It’s never as simple as just focusing on one thing is it? As I get older, I find responsibilities and expectations only increase. As I made a promise to myself to lead a more balanced life when 2013 started, the weekends when I wasn’t working often turned into time to catch up on “the rest of life”.

At times, it felt like I was juggling so many things that I was turning into an octopus. Dealing with those demands weren’t all bad though. I found that while allocating more time to friends, family, and even ‘me’ time required a lot of will power, doing those things made me much happier.

So as 2013 came to an end, I was grateful to get a chance to take some time off and reflect on my year.

What separated my 2013 from 2012 was:
1. I went to bed earlier (not quite the same as 1, but the same idea)
2. I exercised a lot more (at least 2-3 times a week, usually with a 20 minute 5k on the treadmill every week)
3. With the focus on balance, I finally learned to delegate more, and focus on the things that just I can do
4. I found a passion project outside of work, the Chinese American social fraternity I joined in 2010, FF.
5. I dedicated a lot more time to family and made plans for a long overdue trip to China to visit my 90 year old grandmother
6. I paid more attention to the needs of my girlfriend
7. Started reconnecting with some people I’d lost touch with, even on weekdays
8. Started reading using my iPad

These changes have all made significant positive impacts on my life. And I think it’s no coincidence that the Inc article mentions all of them.

Most importantly, I was able to do all of these things, because I finally committed to a philosophy of “And” instead of a philosophy of “Or”. Before 2013, I often viewed options given to me as “this or that”. But once I made balance my top priority, I found that when faced with choices, I started asking myself “how can I do this AND that?”

By asking that question, I find I’ve been able to accomplish more, be happier, without spending a lot more time and energy than before.

In 2014, I hope to incorporate even more ANDs into my life. Two of those ANDs I’ve already started on are Yoga for meditation and being more grateful. I want to incorporate the remaining items in this Inc article to my schedule.

Thank you 2013

What I am thankful for about 2013
- The continue good health of my family
- The love and support of my family and friends
- The many new and wonderful people and experiences that FF brought into my life
- The lessons that I continue to learn each day as an entrepreneur
- How all of it makes me excited about 2014

A survey of the most popular food & drinks apps on iTunes

Last week, we had a product planning session at Gojee. To prepare for it, I went through a comparable analysis of our competitors in the Food & Drinks section of the App Store.

Comparable Analysis is a type of spreadsheet analysis that we did often in my banking days. Each row represents a company, and each column a data point that you want to compare between a set of companies (or ‘comps’).

The trick to a good comparable analysis is in selecting the right comps. For Gojee, I had to find good food apps. So I selected the top 5 iPhone and iPad apps in each of Apple’s store rankings (by Free Downloads, by Paid Downloads, by Grossing $s). I added in a few other apps based on my team’s suggestion for a total of 32 comps.

The whole exercise involved:

  1. selecting the apps to analyze
  2. downloading them
  3. playing with them
  4. tallying the data points in a Google Doc

The whole exercise took about 6 hours, but was tremendously helpful in giving the entire team some insights on where we stood against our competitors.

What We Learned


1. People consume food content in a variety of ways.  Top 5 rankings apps are all very different.  Some ways apps differentiate themselves:

  • Source of Content: Original Content, User Generated Content, Aggregators, Hybrid
  • Media: Video, Pictures, Text
  • Monetization: Paid App, Subscription, In-App Purchases, Ads
  • Cooking Process Focus: Recipe Discovery, Recipe Selection, Recipe Preparation/Cooking
  • Design: Raw, Functional Focus, Style Focus

2. Original content and UGC apps are most common in the space:

  • 16/32 had original content (Food Network)
  • 10/32 had user-generated content (Allrecipes)
  • 4/32 are aggregators (Gojee & Yummly)
  • the rest were a hybrid.

3. 75%+ of the top apps had tools to help the user through the full cooking process (discovery, selection, preparation)

4. Quality/Convenience over Quantity
There was no correlation between number of recipes an app had and its rankings in the store.  Only a few had over 10k+ recipes.  Allrecipes differentiated itself by having the most recipes.  No one tries to claim number 2.

5. Only 2/32 apps required logging in

6. Only 2/32 apps had some form of personalization, in general apps were used as a general source of information.

1. The top rankings apps competed heavily on different ways of helping people in the recipe discovery process, with varying functionalities tied to the preparation process

2. Most common way free apps monetized was through ads.  Most common way paid apps monetized was through in-app payment.

3. Out of apps with high rankings in the store, only Jamie Oliver’s app has videos (how-tos).

1. The highest grossing iPad apps are all magazines, NOT paid apps

2. The most common way free apps monetized was through in-app purchases of subscriptions, content, or another app.

3. Fully functional
Top ranking apps by downloads (paid or free) all had functionalities that made them Kitchen companions that covered the full spectrum of the cooking process (recipe discovery, selection, and meal preparation)


As for how all of this applies to Gojee, we found we are one of the strongest apps for recipe discovery and general design.  We also have work to do on the recipe selection and meal preparation process.

The comparisons also allowed us to evaluate how video is being leveraged in food apps. It is an area that we feel is the future as LTE proliferates.  The lack of a dominant food video app in the space is a sign of opportunity and a void that needs to be filled.

Having a disciplined method to evaluate the competition is crucial to building a successful product in a crowded market place.

If you are building a product and want to do your own comparable product analysis, you can take a look at the sample google doc I used here.



‘Linsanity’ is a Linstant Classic – What it was like seeing it premier night and the day after


I saw Linsanity last night on it’s premier night in New York.  It is an inspirational and thought-provoking piece.  And I highly recommend everyone to make time to go see it.  It will make your day and leave you feeling ready to take on the world.


The Review


The story that the film tells is nothing unique, it’s a classic underdog story.  What makes it refreshing is that in an age of hero-worship by the media, this film does the opposite.  It gives an introspective view of what ultimately brought Jeremy success–the fundamentals of boring old-fashioned values of Religion, Family, Work Ethic, and Perseverance.


Thoughts from the Film


What got Jeremy Lin to where he is today, wasn’t just God-given talent and hard work.  He had to fight hard every step of the way.  When he faltered it was his beliefs and his family that gave him the strength to go on.


Taking a step back, the movie almost is more of a tribute to everything but him.  The timing of things, his family, his coaches, and circumstances outside of his control.  The core message I took away from the movie: Behind every success story there is the best supporting cast in the world.  We should all spend a little less time focusing on ourselves and more on others around us.


Enough reflection, some recollections of the night!


The Premier Night Experience
Out of the 150 seat theater at least 135 was filled for the 10:30 showing.  There was a buzz and excitement before the movie even started.


Some memorable moments:
  • The first scene that Jeremy first shows up, everyone in the audienced clapped
  • When Jeremy played the same songs three years in a row and explained he turned in the first two tests in Harvard blank, everyone laughed and saw a human side to him.
  • Nobody seemed impressed with his 800 SAT math score.


Personally, it connected with me on so many levels, as a first-generation Asian living and growing up in America, as an entrepreneur trying to make it against all odds (though it all seems less daunting now compared to what Jeremy went through…), and as an NY basketballer  scraping around for game every weekend for the past 7 years.


While watching the film, all the sensations and pride I felt of being an often-overlooked and overly-stereotyped Asian-American male during Linsanity flooded back to me.


Supporting the Film


Right now the film is going through its own underdog experience as it is in limited release, being tested in multiple markets (LA, NY, Chicago, Houston, San Diego, San Francisco, Washington D.C.).  I really do hope the film goes to broader release, and will do what I can to help.


To start, my social, Chinese fraternity FF (it is the oldest Chinese fraternity in America)  is organizing a private showing for 150 people at the AMC Village.


I also created this FB page to spread the word, please like it and share it with anyone you know in NY!


More people should hear the story, and not just Asians!


I am very glad to have watched this movie tonight.  It is by far the most memorable birthday night celebration I’ve had in my 30 years.


#linspired #linstantClassic


Building a successful Internet company is about convenience and one more thing

Read this article about Ev Williams’ interpretation of how to build a successful business on the internet.

A company that helps people do something they already do easier and faster is going to succeed.

The first generation dating sites made it easier for a lot of people to find a date.  The next generation dating sites are increasing the turnover and lowering the time investment people have to make.  Grinder makes finding a one night stand as simple as opening an app.  Coffee meets bagel makes finding a date as simple as connecting your Facebook account, and the dating decision a simple yes/no on both sides.  None of the new dating sites are *more* complicated to use than e-Harmony.

The first generation social networks made it easier for people to connect with their friends.  Even though it takes effort to build a profile and upload photos, it still is way better than keeping track of people through an address book and calling them.  The next generation social networks once again lower the time investment for those users.  Twitter only requires 140 characters.  Instagram just requires you to take a photo.  None of the new social networks are *more* complicated to use than Facebook.

With Gojee, we learned Ev’s point first hand.

Our first idea: a grocery reward card data aggregator was complicated and a new idea.  People would have to form a new behavior.  The engineering was complicated because no body had yet invented the wheel in the space.  An idea like that required heavy up-front investment and a long runway, something a first time founding team has a hard time doing.

Our next idea: make recipe finding easier and more pleasurable.  It appealed to the most basic human needs of wanting to be fed and be surrounded by beautiful things.  As soon as it launched, people took noticed and started using it.

Even though Twitter, Coffee Meets Bagel and Gojee have all made our users lives more convenient, our success are vastly different.  The key drivers of success come down to the size and monetizability of the user groups whose lives we make better.

Successful business either needs to be able to convince a few people to fork over a lot of money, or a lot of people to fork over a few dollars.  Dating sites fall in the first category.  Social networks fall in the second category.

Food sites, they fall somewhere in between.  While there are a lot of people who eat food, a much smaller portion actively use the internet for cooking.  Those who do, certainly don’t pay up for it.  So when you have an idea that definitely improves people’s lives, make sure to ask yourself the next question: “how many people’s lives is this product really helping?  if it’s not 1/10 of the world, then how much are those people willing to pay up for this service?”

On perseverance

“I do not think that there is any other quality so essential to success of any kind as the quality of perseverance. It overcomes almost everything, even nature.”

- John D. Rockefeller

My first product lesson: frequency is everything

I was on a call with a couple of startup founders the other day.  They are just starting and looking for some feedback.  They had a polished pitch.  The idea of special private events seemed exciting, but something about it didn’t sit well with me.  I couldn’t pinpoint it right away, but eventually it came to me.

“How often will people use your product?”, I asked them.

The answer was clear to all of us, not very often.

It brought back memories of when Gojee first started.  We built a website that helped people make better decisions whenever they grocery shopped.  Our idea made logical sense, give people a report card telling them how healthy their shopping trip was, and overtime our users will purchase healthier products and be happier.

There was just one problem with that idea and the special events idea pitched by the founders — we all overlooked the frequency factor.  A truly engaging product, almost always has high frequency of use.  Only then can it take on an addictive element and become a habit for a user.

Facebook, twitter, foursquare, any media sites thrives on frequency.  It was one of the first product lessons we learned when we started Gojee and it is one we repeatedly ask ourselves.

Are we building something that will make a user come back in a short period of time?

The only exception to this question is if infrequent use is just part of the existing norm.  For example, vacation rentals, airline bookings are infrequent but nature.  But maybe the next disruptive player in these industries are startups that figures out how to increase the frequency of use for travel and rentals.

I love Marissa Mayer’s decision

Yes, telecommuting does have its place in today’s world.  Yes, it does make employees happier.  But what if those happy employees aren’t the ones that are right for your company?

It all comes down to culture and execution.  

1) How important is culture for the success of your company?  (It matters a lot for a tech company like Gojee or a company in transition like Yahoo, it matters less for a call center or a manufacturing job)

2) What constitutes successful execution for your company?  (At a tech company it’s the ideas that teams generate and their successful implementation.  At a call center, it’s the service representative answering a call courteously and helpfully).

It’s pretty clear which type of companies benefit from telecommuting.  For organizations where teamwork is paramount to success, face-to-face interaction is still better than Google Hangout, Skype, and emails.

So sorry mr. strategy consultant, I don’t buy that strategy and leadership will take care of everything.  That’s only true if execution is guaranteed.  And we both know that execution of anything worth doing is anything but guaranteed.  Reacting to the surprises along the way is much easier done in person.

We’d kid ourselves to think, that operational deficiencies isn’t costly and directly affect the ability of leaders to strategize and set direction for a team of any size.  A misunderstanding over email, a lack of transparency on progress, or a bad internet connection that interrupts a meeting can have huge impact on team chemistry and the success of a project.

And sorry Mrs. Francke.  This isn’t great for parents who want to spend all day with their kids.  But perhaps Yahoo or any company looking for guaranteed execution requires people with a different set of priorities.  What happens if you have an important meeting and your kids are crying in the background?  What takes priority?  Coming from a background where my parents made many sacrifices for me, it’s a tough decision. Maybe by working from the office, that decision with no right answer doesn’t have to be made.

And for every model worker there is one of these.  Maybe it’s actually easier to build true trust in a physical environment, because it builds camaraderie and human interaction.  And maybe workers can actually be happier because they see proof every day that they are part of something bigger than themselves.

I’m not saying every job at every company requires perfect attendance.  I’m saying that there are significant benefits to the physical work place, especially for certain jobs and industries.

In Yahoo’s case, it can be a huge difference maker, but the management team will need a good plan for what the workplace culture and identity will be from now on.  And if so, maybe those who decide to stay will actually end up happier and more productive than before.