Tuesday afternoon scanning the business headlines I saw Markus Frind has decided to sell Plenty of Fish to People Media Group, Match.com's parent company. The sell price is a reported $575 million of which Markus gets to keep everything after lawyers and accountant fees. What a great story that should serve to inspire developers everywhere.
I realize many of you may not know who Markus is or what Plenty of Fish is for that matter. In 2003 Markus Frind was your typical developer moving from contract position to contract position. He wanted to learn ASP.NET, I assume because those positions paid better than classic ASP or PHP, so he decided to create an application to both learn some skills as well as prove he knew them for interviews. After some conversations with friends he decided to create a free online dating application. It was not long after that the site began to get real traction and by 2004 Plenty of Fish was rocketing to the top of the most visited web sites list.
I started following Markus somewhere in the 2006 time frame when Jeremy Schoemaker (aka Shoe Money) called Markus out as a fraud. Jeremy has his on rags to riches story and has profited by sharing the details in an effort to help others do the same. At the time Jeremy had a podcast focusing on Internet marketing called Net Income and interview Markus to challenge some claims he had recently made.
You see Markus was not running LAMP (Linux, Apache, MySQL and PHP) but rather ASP.NET, IIS and SQL Server. Not only that he was running a high traffic site with just 4 physical servers. One web server, one image server, one e-mail server and one database server. After news started circulating that Markus was running this online dating giant using ASP.NET and 4 physical servers Jeremy and other LAMP stack devotes called shenanigans. Part of the ridicule targeted the Microsoft stack, the other part came from running so few servers.
The other aspect many have a difficult time grokking is how can one man accomplish so much? The reality is most sites and applications are relatively simple. I have been inside large companies with 'large' web sites and I have done things on my own and worked with small businesses. The reality is the applications are very similar in structure. For example a typical e-commerce site has less than 10 pages driving the experience. The back-end can be very complex, but again it does not need to be. My impression of large companies is they tend to over engineer and hire too many developers.
For example I was once on a team of 45 building an application. The application itself was not that complex, but what made things complex were the 45 team members. This created overhead and bureaucracy. Neither of these features as value, but instead drag down productivity and add overhead costs to the product. On this project around 50% of my time was spent dealing with meetings and other paperwork that added no value to the process and the meetings never really improved what I was developing.
Another team had 16 developers. We spent 25 hours a week in meetings. Again no value was added. After 6 months the team accomplished little tangible value. I felt held back the entire time. Tasks the team determined would take 2 weeks took me 2 hours, etc.
I share these personal experiences to demonstrate the way enterprises typically bog themselves down to a pit of unproductivity. Compare that to my experiences as a one man team working for small business and startups. The productivity velocity was far better. What seems to take 6 months in a typical enterprise takes a week or two in the one man scenario. In short it is easy to be decisive when you are on your own. You also do not need to worry about writing code and managing the code writing process so other developers can follow along. You are also free to fail fast, meaning you can try things out, see if they work and change things rapidly. There is no 6 month deployment cycle to worry about.
Markus benefited by staying a company of one for many years. Here he was a one man band generating more traffic than larger competition, like Match.com and eHarmony.com. To me he proved several things, it does not take hundreds of servers nor large technical teams to accomplish large scale success. On top of that he did it before the cloud was available.
I pulled this 2006 quote from his Blog:
"I spent every waking minute when I wasn't at my day job reading, studying, and learning. I picked out "enemies" and did everything I could to defeat them which ment being bigger then them. I refused to accept defeat of any kind, and I constantly forced myself to test new things. I never tried to perfect anything it didn't matter if things didn't work 100% as long as it was good enough I would move onto the next thing." Inc. 2009
In other words he worked his butt off and focused on his business. He learned not only the development stack, but online and affiliate marketing. The latter is how you get money. As you read through his Blog you also realize he spent a lot of time learning the online dating business. This allowed him to make many innovations that set Plenty of Fish apart from the competition.
In a 2006 Net Income interview Markus made a comment that has stuck with me over the past 9 years. It goes something like this, "most places don't take the time to code things right, instead they just throw more hardware at it trying to solve problems". Again let me pull a quote from the Plenty of Fish Blog about the hardware Markus used to build his community:
"The site(plentyoffish.com and Forums.plentyoffish.com) serves ~500 million pageviews a month and does so using 1 DB server and 1 Web Server which is a far cry from the industry standard of 300+ servers for a site of this size."
He upgraded to a second web server in 2007. This is where I learned there is a 64,000 connection limit per IP address. Earlier in 2007 Markus talks about the actual server he used. Nothing spectacular, just something like the average company could easily afford. But what he proves is it does not take a lot of iron and CPU to run a large scale web site, even in 2007. Today I would use Azure web sites with other Azure features to cost effectively run my infrastructure. Based on some of the traffic stats Markus shares I think his monthly 2007 costs would be roughly $3-4000 for storage, traffic and web site. Not bad considering PlentyOfFish.com's monthly traffic and revenues.
"Most of the largest dating sites are spending over 1 million a month on hardware and tech staff to keep it running. The only reason my site is still in business and works so well is because i’ve found a way to solve the technical side of the problem." November 2006
In April of 2007 Markus did a Channel 9 interview where he talks about the company's background and the server infrastructure it used. I love this statement in the interview description:
"The site runs on Windows Server 2003, Standard Edition, 31 million page pages - 2 million page views per hour at peak, 40-50K concurrent users – on ONE IIS 6 server at 65% capacity. "
At the time I was hosting hundreds of customer site, on a single web server. My load was not near as high as Plenty of Fish's but my experiences echo his. My usage was around 5% of his, but again I saw very little server capacity impact. I experienced several traffic spikes when customers were featured on television, radio or did an advertising blitz. Again even those spikes were easy to handle.
And yes he did it by himself. Another quote from his Blog:
"It amazes me after all this time I have still ran the company as the solo shareholder/solo employee for longer than i’ve run it with employees. I got to 10 million + users without hiring a single employee."
Despite publishing several posts on his Blog about wanting to hire developers he did not pull the trigger till May 2009 when he hired his first programmer in May 2009 and a DBA later. Today the stories of the Plenty of Fish acquisition say the company has 75 employees. I don't know how many developers and IT admins that includes. My suspicions are very few. The majority of his staff is most likely customer support, you know dealing with customer questions and issues.
Markus and Plenty of Fish have long been a testament and inspiration to me over the years. I regret not being able to focus enough to accomplish something similar. His use of lean web development has influenced my approach. Not only has Plenty of Fish proven the lean approach works, many other companies have as well.
Back to the Net Income interview. There is a question by a Markus Frind fanboy about being a one man show and is that a disadvantage when trying to sell. I encourage you to listen till the end for this question, you may find the called somewhat familiar :).
I am going to finish this post with another quote from the 2006 Net Income interview. "It's so easy to make money online once you know what you are doing, there is no incentive to sell." He goes on to say he might sell in the distant future. Well that time has finally come and Markus is cashing in and riding off into the sunset. Thanks for being an inspiration both technically and as an entrepreneur.
Here are some links to various Markus and Plenty of Fish content I used in this article: