[This is the second part of my perspectives on Soccermetrics on its 5th anniversary.]

While I was expanding from writing about other people’s research in soccer analytics to conducting my own, events were occurring on a backdrop of a Soccermetrics’ transition from a blog to something more formal.  I attended the MIT Sloan Sports Analytics Conference in 2010 believing that sports analytics could become really big and left convinced of it.  In 2011 the field had exploded with soccer analytics front and center of the phenomenon.

It was then that I made two fateful decisions: convert Soccermetrics into a company and relocate to Atlanta.  I moved to Atlanta — and not the Bay Area — because of the low cost of living, proximity to international travel, and overall familiarity with the growing technology scene.  I’m from the southeastern US, I’m a Georgia Tech alumnus, and I like living here, so launching the company allowed me to live in a place where I wanted to be.

I do not regret the decision to recreate Soccermetrics as a tech startup company.  I had the opportunity to return to a city that I love and a part of the country where my family resides.  In business, one learns a lot of things by doing.  I learned a lot of things about myself — skills I never knew I had, and flaws, deficiencies, and fears that I have had to face up to.

I established strong contacts with the Atlanta technology startup community that I value on a personal and professional level.  I was thrilled and humbled at how many important people in Atlanta technology circles were regular readers of Soccermetrics. The power of serendipitous encounters is amazing.

I was blessed to have a great advisory board of researchers, business people, and later investors who provided valuable advice and mentorship.  Matt Stigall answered my call for a business partner even though he never saw my ad (talk about a serendipitous encounter!) and Aaron Nielsen brought his impressive database to the company.

I made a lot of valuable contacts and had engaging conversations with people throughout the football industry, some of whom were quite well-known, about their use of data, their perspective on analytics in the sport, and what their needs or desires were. London became something like a second home to me.

Along with the rest of my software development and data intelligence teams, we built some useful and valuable software that is the foundation of our analytics work. Soccermetrics grew its knowledge base through its Newsletter and Podcast resources, which allowed me to present  conversations with some leading names in soccer business and analytics research.

A lot of people fell over themselves to get my attention, but it’s too bad that almost all of them were either service providers who I couldn’t afford and didn’t need at this stage no matter how much they tried to convince me otherwise, or job seekers with CVs attached eager to join in any capacity possible, even though I didn’t know how I was going to keep paying for myself much less someone else.  I will never forget the day when Matt posted a job notice  in the morning on Reddit and we received almost 4,000 visits and 80 resumes by nightfall.  Even today I receive 3-5 resumes in a week.

I have always respected entrepreneurs, but I have even greater respect for those who have built multiple ventures, some successfully, others not so much, but always boldly attempting to cause change in the world.  It takes big brass attachments to start your own business, no matter what it is.  Yet it’s not enough to do something you like; it’s essential to do something that addresses a real human need and enough people are willing to pay for.  And from this hindsight view, I don’t believe that Soccermetrics was suited to be a startup company when I started it.

There are many small businesses and consultancies that are successful, sustainable, and profitable for their proprietors, yet are not scalable ventures.  There’s nothing wrong with that, but such businesses will never attract risk capital from a VC (and it would be foolish for those owners to take that kind of money anyway).  Scalable ventures require a large and growing addressable market, a repeatable business model, high profit margins, and minimal incremental costs.

Football clubs — and all professional sport clubs — have the illusion of a large market because of the billions of dollars outlaid to players and technical staff, but the true addressable market size, the money available to spend on products like yours, is very small.

The time and money required to acquire a sports customer is extremely long and expensive, and in the case of football analytics as I practice it, not scalable.  It is very easy to have someone interested in a blog post on some topic, yet extremely difficult to convince the same person that they are in need of your service and should hire you.

The downward pressure on fees is very strong because the ultimate competitors are the interns who get paid little or the passionate fan who sends his own customized analysis to his favorite team.  Furthermore the costs are enormous and take different forms; a firm devoted exclusively to data analysis must first purchase expensive data feeds, and a firm that performs data collection and analysis for clubs must pay for human-led collection and validation of match data and the software used to enable them.

It is essential that the product one builds is something that users must have; a product that is “cool to have” is not enough!   And in the professional sports industry, unless one is offering equipment for the playing staff or operational software for the front office (e.g. accounting, ticketing, CRM, basically enterprise software), everything else is “cool to have”.

Perhaps the most important lesson I have learned is that in the early stages of a startup, new technology development stops.  There is simply no time to develop it while attempting to find product/market fit.  What that means is that your core technology had best be mature enough to create a salable product before you consider starting a company around it.  And soccer analytics research is not even close to being mature.

Well the world doesn’t stand still in two years, and soccer analytics are no exception.  I’ll discuss that in part three.