I had an entrepreneurial itch and I was dying to scratch it. I also knew - when it came to business - I knew nothing.

I decided to begin two synergistic projects - I would obtain an MBA and start an independent game studio.

I started this project by establishing an Agile process that would carry me from the beginning to the end. I created an iterative process with 2-week Sprints.

To begin, I wrote a story about the experience a user would have with my product at the end of the Sprint. I broke this down into requirements.

At the end of every cycle - I did a post-mortem. I asked myself, 'Are the requirements fulfilled? Did I succeed? Why or why not?'. I reviewed the twelve agile tenets and asked myself, 'Am I being Agile? How can I be more Agile?'.

I decided on any alterations. Then I wrote the next Sprint's story.

I was not short on game ideas. Never have been! I'd been participating in game developer communities for many years and had seen firsthand that a great game idea, even great execution, was seldom successful. In fact, the one success story I had seen firsthand had done neither of those things. What they had done - was find an underserved market.

I decided to do the same.

My hypothesis : Quality would carry me up the App Store rankings to capture the market.

Spoiler : I was wrong.

I scoured the Android and Apple app stores looking for inspiration. I played a gazillion mobile games. But I came up empty. Inspirado wouldn't find me until a sometimes freelance client reached out about a new project - could I make a Tangram game for Kinect in two weeks?

So I researched. The project did not go forward but I had found a project that fit my criteria.

The top versions on the respective marketplaces had over 1 million downloads. I found a blog post from the Apple version where the developer said he had quit his job and was overwhelmed by his apps success.

Even better - they were not the same app! This indicated a low technical capacity. I knew I could surpass their usability and art direction. I knew I could have a bigger audience. I dived deep on Tangram - and I came to understand - Tangram is forever.

Tangram went global by boat and baggage train in the 1800s. It seems to have originated in China and spread quickly from there. By , Tangram was a global phenomenon. Tangram books were published in every language in every country.

The great 19th century French artist, was so enamored he created his own book! His beautifully rendered book became my muse.

Maybe you recognize the pieces. Perhaps with a smile. It is hard to not feel wonder when these 7 simple shapes form nearly any object.

My game design was simple - each completed puzzle unlocks a new object in an unfolding scene. This provided an immediate reward towards a long term one. New levels unlocked at intervals different than level completion. Later, I would add new puzzle pieces as yet another staggered unlockable.

I had built User Personas but I never built Buyer Personas. This locked my mindset into an ad-driven revenue. In retrospect, this is was my greatest failure on the project.

I relied on qualitative results for my Primary Research. Quantitative tracking would have prevented problems down the road. Looking back, my design reflects the problems I would have with my first time user funnel but in the moment - I thought everything was peachy.

My first MVP ran on desktop using the mouse. I quickly moved to validate on device. I obtained the worst Windows Phone possible. For development I have a worst device policy. Using the best device obscures performance issues.

On device invalidated two aspects of my desktop development.

The UI suffered because it did not compensate for vision being blocked by the finger.

The algorithm I used to determine puzzle completion used pixel counting. This was non-performant on device. I had to develop a method of determining overlap of triangles using integrals.

I alpha tested my game with over a hundred in-person users. I was working on my MBA, volunteering at a home school co-op, and frequenting coffee shops. This gave me access to users of all ages, backgrounds, and technology familiarity.

I documented a near 100% completion rate for pre-K and retirees.

I like to enter contests. They provide an educated, objective viewpoint that is difficult to surface in face-to-face interactions. I decided to enter Microsoft's Next App Star contest.

To make it to the deadline I had to choose a subset of features to implement. I created a tutorial and two levels : one fish themed, the other Egypt themed.

As a top 16 finalist, my entry netted me award money and other prizes. It was also a huge marketing boost. That I was completely unprepared for.

The first version of OMG tracked one custom analytic - puzzle completions.

I chose the wrong technology. I bet on the wrong horse. With all the data in front of me - and a goal of cross-platform domination - I had picked

First time user funnel completion was the one metric I cared about. Losing players early is a lose-lose. They don't even know if they like the game yet!

I wanted players to experience the game before they decided to quit.

I found problems, theorized solutions and slowly improved the FTUE funnel. The initial steps that didn't work focused on telling. The new high completion steps focus on tell.

Show - don't tell. Show and tell if you must but as the following story demonstrates - don't count on it to help.

The first time user funnel was looking nearly perfect. Better than perfect. Perfect was a 95% pass rate - 5% error is acceptable. 100% in a day was not uncommon for most funnel steps.

But I still had a flipping problem. A literal flippin' problem. Double tap as a gesture was a disaster - the beta FTUE success rate had been below 50%. The most recent version could dip down to the 70s.

Iteration had not fixed it - it had slowed the leak but the leak was still there. Losing 30% of users was a hemmorhage.

Show what competition used.

Show translation problem.

Show slower animation. (didn't try)

Add extra piece. Reduces elegance.

I had hypothesized that quality would lead me to the top of the charts. While I had steadily climbed, my growth had stagnated. Even then - my belief had not.

Then one day, suddenly, worldwide, there was a new number one tangram app. Within a week it had garnered 1000 unanimous 5-star reviews. This is not actually possible. This was paid for by a marketing firm. Over time, it's real rating would reveal itself, but it's market position remains secure.

I had to face it - my hypothesis was incorrect. To achieve my goal in a mature marketplace - I needed mature marketing. A level of marketing beyond my skill and means. It was time to stop.