The Year My Startup Broke (in a good way)
December 2009 marked five years of operation for FanFooty, my Australian fantasy football site. I guess after five years you can't call it a start-up any more, it's just a business. Time is as good as any to look for backward and forward, now that I have reached what I consider to be a tipping point.
Back in December of 2004, my new housemate Tai Tran came to me and said that if I had an idea for a business to start, then he'd be interested in doing it together. Tai was about my age, had a professional background in corporate development, and was a likable bloke. He had the coding chops and I had the writing ability, and we both had proven track records in our fields. It seemed like a workable pairing.
I was introduced to the fantasy sports concept by some American friends of mine and got hooked. Looking around the Australian scene, I could see that it was sorely underdeveloped compared to the American and even the English industries. Even at that stage it was a billion-dollar business in the US involving 1 in 12 Americans, with ESPN, Yahoo, Sportsline.com and a host of others competing for big dollars; plus every major newspaper in the UK had a fantasy EPL game, with prizes well into six figure pound ranges. At the time here, there was only the AFL Dream Team competition with less than 50,000 players, and a handful of amateur efforts with virtually no patronage, but the upcoming year of 2005 was when DT numbers exploded to over 130,000 and the graph has been jumping every year since. It was the right time to start the business.
As with countless other startups in a myriad industries, our feeble plan as to what our business would look like and how it would make its money did not survive engagement with reality. My original thought was that the Australian industry had developed too far along English lines, focused around mass-entry salary cap competitions, and there was a perfect opportunity to expand into the American-style private draft leagues for small groups of 8-16 friends, and we would be at the forefront of it. The first thing we found was how difficult it was to build a private draft fantasy football application. I was learning PHP from scratch, not having had any programming education since high school (though I taught myself BASIC on my old Commodore 128D), but I had the business logic all in my head so it became rather time-consuming to explain how the application worked to Tai, who had no knowledge of or interest in sport, so that he could go away and write the code.
Nevertheless, we got the code done with Tai learning as much about AJAX techniques for the very tricky live drafting component as I did about PHP for the less difficult pages, and we had a workable application. Then came the dose of reality. We had very few customers. With no venture capital behind us to fund any sort of marketing budget, all I could do was hit the message boards to talk it up. That wasn't enough. That part of the business has been a waste of time, to put it bluntly. Five years later and no one, including VirtualSports as the official provider of its version called Premium Dream Team, has made much of a fist of the US-style private draft form of the game. The market opportunity still exists.
Back in FanFooty's debut AFL season of 2005, it didn't take me long to figure out a second possible revenue stream: live fantasy scoring. The official Dream Team site wasn't offering it at the time, but it seemed like a no-brainer for me as I was used to Sportsline and the other American fantasy providers who all provided live scoring. The feature was moderately popular in the first year, with site traffic reaching the dizzy heights of 20,000 page views in a couple of weeks and even swamping our meagre server resources by the start of May. In those first few months I created more and more pages from a combination of my own inspiration and feedback from users, setting up some of the fundamental features of the site which still drive traffic to this day. Even with Google AdSense, nevertheless, the revenue was barely covering the site hosting expenses.
It was 2006 - the first year of the Herald Sun Super Coach competition - when the site developed serious growing pains, and they hit very hard. At this stage we were still hosted on a shared server but we ended up shutting down the entire box for hours at a time during games on a weekend, due to a combination of a quadrupling of raw demand plus some loose code. It got bad enough that I decided I had to institute paid memberships to slow the torrent of users. That ended up being another big mistake. Barely 80 lots of $5/$10 later, I backed down. There was no other choice, Tai and I had to figure out how to build a more scaleable system or we wouldn't have a viable business. In the mean time, the site was developing an unwanted reputation for being unreachable on weekends and dropping connectivity at the drop of a hat, with "is fanfooty down?" an all too common refrain on the message boards. We were just too popular.
Thus it was that we started developing the techniques that have made FanFooty different to most other sites. It's not something that you can see - indeed, FanFooty's design aesthetic is rather minimalist, even old-fashioned in some respects. Over time, we did some benchmarks and watched the health or otherwise of the live scoring section of the site under different traffic conditions. We eventually decided to abandon the LAMP stack entirely. Our new architecture involved a second server running lighttpd instead of Apache, with no database or scripting languages installed. This allowed maximum scalability for a small subset of pages that was driving the vast majority of our page views. Purist pro programmer friends scoffed at the client-heavy AJAX scripting, but it worked for us with our limited server resource budget.
By this stage Tai and I were both gaining confidence in our abilities, and were enjoying the learning process of discovering new bits of code to use as weapons in the neverending war against our own ignorance. That's not to say that all was clean sailing. I got very angry at times during server downtimes, and I am ashamed to say that I took some of it out on poor Tai. For his part, Tai did some very good work but I was the one providing just about all of the creativity on the project, which caused further tensions. In addition, my sleeping schedule was occasionally moving around the clock, so that for weeks at a time our waking times weren't intersecting all that often.
Nevertheless, 2007 was a good year on many fronts. I was a big fan of the Hitwise model of startups, insomuch as you have a cash cow business and a home run business operating side by side: the former funding the latter, and the latter eventually being the one that takes you over the top. For our purposes, FanFooty was the cash cow and Tinfinger was the home run. Much of 2007 was spent in developing Tinfinger, a "human omnibus" with elements of Wikipedia, TechMeme and Squidoo/Mahalo, plus a social networking app bolted on the side. FanFooty grew in traffic by 640% year-on-year, so it was by no means being neglected, and many more important features were created this year, including the FanFooty blog and the Coaches Box podcast. Tinfinger was where most of our energy was directed, and it was here that the cracks in the partnership between myself and Tai finally opened up. Tai has many fine qualities, but he doesn't possess the spark of originality of thought that I was looking for as part of a good founder's skill set. That's not to say that a founder necessarily needs to be particularly original to create a successful business. I, however, needed someone who was prepared to invest a little more brainpower and take the pressure off me to formulate every little bit of our strategy.
I don't mean to denigrate Tai by saying the above, either, so let me talk a bit about how good it was to have him as a co-founder, and how invaluable he was at the darkest times. There were periods where I needed a friend more than I needed a business partner, and he was there to listen to me as I sloooowly opened up. Similarly for him, he went through some terrible personal stuff around that time and (I hope) I was able to contribute in some small way to his recovery. Often our mood swings would complement each other so that one would be able to share his energy to encourage the other when they were down.
2007 was also a good year for advertising, and with growth to above half a million page impressions per week, advertisers were starting to take notice. Through AdSense we started getting targeted campaigns from major advertisers like Pepsi, Schick and Gillette, which added up to our first year of profit, albeit still nowhere near enough to pay both of us living expenses.
2008 was quite the opposite. Tinfinger launched in January and, unlike FanFooty, we couldn't rely on the large amount of pent-up demand that greeted us back in early 05. The ego-arbitrage space was already well populated by companies larger, smarter and more experienced than us. The division of labour between myself and Tai became more pronounced, as my programming skills rivalled and in some areas surpassed his. After a fair few shouting matches and other unpleasantness, I knocked it on the head in May and negotiated an agreement whereby he would continue to be paid a percentage of FanFooty profits but we would drop Tinfinger. The split wasn't what either of us wanted, which indicated that it was fair at the least.
In addition, we had made another big mistake that year in changing advertising providers, going from AdSense to local firm 3dinteractive (part of ASX-listed Q Ltd). I don't wish to disparage 3di either, suffice it to say that nothing good whatsoever came out of that relationship. Our traffic doubled off a very healthy base, going from half a million to one million page views per week, but we had very little to show for it at the end of the season, having earned precisely zero dollars out of the April-through-July period (for reasons I won't go into). By the time we got back onto AdSense the major advertisers' campaigns had deserted us, and we were lucky to break even across the year.
The start of 2009 looked like being even more disastrous as the AdSense numbers fell into the toilet in January as the GFC hit, but two fortunate events occurred. First, we were headhunted by Platform 9, an ad agency division of ninemsn whose primary focus is on video ads, but also operates a system much like AdSense for remnant display inventory for Australian advertisers. Their numbers took a little while to come on but they have been very pleasing, with the video element providing an invaluable auxiliary revenue stream. Second, we were rather lucky to get our first direct sponsorship deal, with Betfair. Every major football media property aligned itself with one of the betting providers in 2009 after the relaxation of legislation which previously preventing them from advertising. The combination of these two events made 2009 a much more comfortable year in terms of revenue, especially since traffic doubled yet again to average two million page impressions per week.
By this time Tai was doing virtually no work on FanFooty, as per our agreement, but this had been the case for two or three years previous anyway. However, he came to me late in the season to talk about dissolving our previous agreement and freeing both of us up for the future. For this, I can only thank Tai for his maturity and understanding. He has a lovely girl who makes a great partner for him now, and he is over most of the problems that beset him during the worst of our times at the Geelong house, some of which were due to our business.
I have since moved to Brunswick and have settled down into a new lifestyle. I can look forward to a future where I have complete control over all aspects of my life and business, and Tai can do the same with the projects he has in development. I don't agree with those people who look back on a five-year span like this and say something like, "I wouldn't have done anything differently." That's silly. Of course I made mistakes, and I hurt people, and I let people down at times, not least of all myself. I would like to have avoided the decisions I made that wasted some of the precious and limited time I have on this earth to make something of myself.
Nevertheless, I can't change anything now, and I can at least live with what has happened. Not that many people ask me for advice, but if they did, I would tell them to get themselves a time machine, go back five years and start a business. It takes five years to get to the point where you know what you're doing and where you're going.
I am happy to have survived that process, where many of my contemporaries have not. I started at around the same time as the likes of Cameron Reilly, Bronwen Clune, Ben Barren and Duncan Riley, all of whom I like greatly and admire for their best qualities. All but Ben were married when they started, and now none of them are. None of them are still working the same startup as when they started. Web 2.0 has been and gone, and it has left many of my friends with not much at all to show for it despite years of hard yakka.
At this point my old technology journalist mates from my previous life might snigger behind their hands, but most of the jobs that they were in five or ten years ago have disappeared too, into the ether or perhaps only half-replaced by itinerant freelance table scraps. Worthy efforts like Hydrapinion and iTWire are not introducing new journos into the industry from what I can see, and aren't ambitious enough to start generating new jobs.
From the above, some of you might think I have been exceedingly lucky to even get this far, despite the business still not earning enough to give me a decent wage. That is probably true. I have put a lot of hard work into FanFooty over the years, though, and I think I deserve it. I have only survived in part because I am not married or even in a relationship. Only now do I feel it would be fair to subject a girl to my lifestyle. Most would still look down their nose at my lack of financial stability. On that front, things are looking a little more lively in 2010 than at any recent time, so there's a lot to look forward to.
I realise this blog post got a bit long and rambling, so for that I apologise. I might follow up with some further thoughts later in the week. Thanks for reading this far, I'm off to the parents' place for a weekly roast dinner. Some things never change. :)
Back in December of 2004, my new housemate Tai Tran came to me and said that if I had an idea for a business to start, then he'd be interested in doing it together. Tai was about my age, had a professional background in corporate development, and was a likable bloke. He had the coding chops and I had the writing ability, and we both had proven track records in our fields. It seemed like a workable pairing.
I was introduced to the fantasy sports concept by some American friends of mine and got hooked. Looking around the Australian scene, I could see that it was sorely underdeveloped compared to the American and even the English industries. Even at that stage it was a billion-dollar business in the US involving 1 in 12 Americans, with ESPN, Yahoo, Sportsline.com and a host of others competing for big dollars; plus every major newspaper in the UK had a fantasy EPL game, with prizes well into six figure pound ranges. At the time here, there was only the AFL Dream Team competition with less than 50,000 players, and a handful of amateur efforts with virtually no patronage, but the upcoming year of 2005 was when DT numbers exploded to over 130,000 and the graph has been jumping every year since. It was the right time to start the business.
As with countless other startups in a myriad industries, our feeble plan as to what our business would look like and how it would make its money did not survive engagement with reality. My original thought was that the Australian industry had developed too far along English lines, focused around mass-entry salary cap competitions, and there was a perfect opportunity to expand into the American-style private draft leagues for small groups of 8-16 friends, and we would be at the forefront of it. The first thing we found was how difficult it was to build a private draft fantasy football application. I was learning PHP from scratch, not having had any programming education since high school (though I taught myself BASIC on my old Commodore 128D), but I had the business logic all in my head so it became rather time-consuming to explain how the application worked to Tai, who had no knowledge of or interest in sport, so that he could go away and write the code.
Nevertheless, we got the code done with Tai learning as much about AJAX techniques for the very tricky live drafting component as I did about PHP for the less difficult pages, and we had a workable application. Then came the dose of reality. We had very few customers. With no venture capital behind us to fund any sort of marketing budget, all I could do was hit the message boards to talk it up. That wasn't enough. That part of the business has been a waste of time, to put it bluntly. Five years later and no one, including VirtualSports as the official provider of its version called Premium Dream Team, has made much of a fist of the US-style private draft form of the game. The market opportunity still exists.
Back in FanFooty's debut AFL season of 2005, it didn't take me long to figure out a second possible revenue stream: live fantasy scoring. The official Dream Team site wasn't offering it at the time, but it seemed like a no-brainer for me as I was used to Sportsline and the other American fantasy providers who all provided live scoring. The feature was moderately popular in the first year, with site traffic reaching the dizzy heights of 20,000 page views in a couple of weeks and even swamping our meagre server resources by the start of May. In those first few months I created more and more pages from a combination of my own inspiration and feedback from users, setting up some of the fundamental features of the site which still drive traffic to this day. Even with Google AdSense, nevertheless, the revenue was barely covering the site hosting expenses.
It was 2006 - the first year of the Herald Sun Super Coach competition - when the site developed serious growing pains, and they hit very hard. At this stage we were still hosted on a shared server but we ended up shutting down the entire box for hours at a time during games on a weekend, due to a combination of a quadrupling of raw demand plus some loose code. It got bad enough that I decided I had to institute paid memberships to slow the torrent of users. That ended up being another big mistake. Barely 80 lots of $5/$10 later, I backed down. There was no other choice, Tai and I had to figure out how to build a more scaleable system or we wouldn't have a viable business. In the mean time, the site was developing an unwanted reputation for being unreachable on weekends and dropping connectivity at the drop of a hat, with "is fanfooty down?" an all too common refrain on the message boards. We were just too popular.
Thus it was that we started developing the techniques that have made FanFooty different to most other sites. It's not something that you can see - indeed, FanFooty's design aesthetic is rather minimalist, even old-fashioned in some respects. Over time, we did some benchmarks and watched the health or otherwise of the live scoring section of the site under different traffic conditions. We eventually decided to abandon the LAMP stack entirely. Our new architecture involved a second server running lighttpd instead of Apache, with no database or scripting languages installed. This allowed maximum scalability for a small subset of pages that was driving the vast majority of our page views. Purist pro programmer friends scoffed at the client-heavy AJAX scripting, but it worked for us with our limited server resource budget.
By this stage Tai and I were both gaining confidence in our abilities, and were enjoying the learning process of discovering new bits of code to use as weapons in the neverending war against our own ignorance. That's not to say that all was clean sailing. I got very angry at times during server downtimes, and I am ashamed to say that I took some of it out on poor Tai. For his part, Tai did some very good work but I was the one providing just about all of the creativity on the project, which caused further tensions. In addition, my sleeping schedule was occasionally moving around the clock, so that for weeks at a time our waking times weren't intersecting all that often.
Nevertheless, 2007 was a good year on many fronts. I was a big fan of the Hitwise model of startups, insomuch as you have a cash cow business and a home run business operating side by side: the former funding the latter, and the latter eventually being the one that takes you over the top. For our purposes, FanFooty was the cash cow and Tinfinger was the home run. Much of 2007 was spent in developing Tinfinger, a "human omnibus" with elements of Wikipedia, TechMeme and Squidoo/Mahalo, plus a social networking app bolted on the side. FanFooty grew in traffic by 640% year-on-year, so it was by no means being neglected, and many more important features were created this year, including the FanFooty blog and the Coaches Box podcast. Tinfinger was where most of our energy was directed, and it was here that the cracks in the partnership between myself and Tai finally opened up. Tai has many fine qualities, but he doesn't possess the spark of originality of thought that I was looking for as part of a good founder's skill set. That's not to say that a founder necessarily needs to be particularly original to create a successful business. I, however, needed someone who was prepared to invest a little more brainpower and take the pressure off me to formulate every little bit of our strategy.
I don't mean to denigrate Tai by saying the above, either, so let me talk a bit about how good it was to have him as a co-founder, and how invaluable he was at the darkest times. There were periods where I needed a friend more than I needed a business partner, and he was there to listen to me as I sloooowly opened up. Similarly for him, he went through some terrible personal stuff around that time and (I hope) I was able to contribute in some small way to his recovery. Often our mood swings would complement each other so that one would be able to share his energy to encourage the other when they were down.
2007 was also a good year for advertising, and with growth to above half a million page impressions per week, advertisers were starting to take notice. Through AdSense we started getting targeted campaigns from major advertisers like Pepsi, Schick and Gillette, which added up to our first year of profit, albeit still nowhere near enough to pay both of us living expenses.
2008 was quite the opposite. Tinfinger launched in January and, unlike FanFooty, we couldn't rely on the large amount of pent-up demand that greeted us back in early 05. The ego-arbitrage space was already well populated by companies larger, smarter and more experienced than us. The division of labour between myself and Tai became more pronounced, as my programming skills rivalled and in some areas surpassed his. After a fair few shouting matches and other unpleasantness, I knocked it on the head in May and negotiated an agreement whereby he would continue to be paid a percentage of FanFooty profits but we would drop Tinfinger. The split wasn't what either of us wanted, which indicated that it was fair at the least.
In addition, we had made another big mistake that year in changing advertising providers, going from AdSense to local firm 3dinteractive (part of ASX-listed Q Ltd). I don't wish to disparage 3di either, suffice it to say that nothing good whatsoever came out of that relationship. Our traffic doubled off a very healthy base, going from half a million to one million page views per week, but we had very little to show for it at the end of the season, having earned precisely zero dollars out of the April-through-July period (for reasons I won't go into). By the time we got back onto AdSense the major advertisers' campaigns had deserted us, and we were lucky to break even across the year.
The start of 2009 looked like being even more disastrous as the AdSense numbers fell into the toilet in January as the GFC hit, but two fortunate events occurred. First, we were headhunted by Platform 9, an ad agency division of ninemsn whose primary focus is on video ads, but also operates a system much like AdSense for remnant display inventory for Australian advertisers. Their numbers took a little while to come on but they have been very pleasing, with the video element providing an invaluable auxiliary revenue stream. Second, we were rather lucky to get our first direct sponsorship deal, with Betfair. Every major football media property aligned itself with one of the betting providers in 2009 after the relaxation of legislation which previously preventing them from advertising. The combination of these two events made 2009 a much more comfortable year in terms of revenue, especially since traffic doubled yet again to average two million page impressions per week.
By this time Tai was doing virtually no work on FanFooty, as per our agreement, but this had been the case for two or three years previous anyway. However, he came to me late in the season to talk about dissolving our previous agreement and freeing both of us up for the future. For this, I can only thank Tai for his maturity and understanding. He has a lovely girl who makes a great partner for him now, and he is over most of the problems that beset him during the worst of our times at the Geelong house, some of which were due to our business.
I have since moved to Brunswick and have settled down into a new lifestyle. I can look forward to a future where I have complete control over all aspects of my life and business, and Tai can do the same with the projects he has in development. I don't agree with those people who look back on a five-year span like this and say something like, "I wouldn't have done anything differently." That's silly. Of course I made mistakes, and I hurt people, and I let people down at times, not least of all myself. I would like to have avoided the decisions I made that wasted some of the precious and limited time I have on this earth to make something of myself.
Nevertheless, I can't change anything now, and I can at least live with what has happened. Not that many people ask me for advice, but if they did, I would tell them to get themselves a time machine, go back five years and start a business. It takes five years to get to the point where you know what you're doing and where you're going.
I am happy to have survived that process, where many of my contemporaries have not. I started at around the same time as the likes of Cameron Reilly, Bronwen Clune, Ben Barren and Duncan Riley, all of whom I like greatly and admire for their best qualities. All but Ben were married when they started, and now none of them are. None of them are still working the same startup as when they started. Web 2.0 has been and gone, and it has left many of my friends with not much at all to show for it despite years of hard yakka.
At this point my old technology journalist mates from my previous life might snigger behind their hands, but most of the jobs that they were in five or ten years ago have disappeared too, into the ether or perhaps only half-replaced by itinerant freelance table scraps. Worthy efforts like Hydrapinion and iTWire are not introducing new journos into the industry from what I can see, and aren't ambitious enough to start generating new jobs.
From the above, some of you might think I have been exceedingly lucky to even get this far, despite the business still not earning enough to give me a decent wage. That is probably true. I have put a lot of hard work into FanFooty over the years, though, and I think I deserve it. I have only survived in part because I am not married or even in a relationship. Only now do I feel it would be fair to subject a girl to my lifestyle. Most would still look down their nose at my lack of financial stability. On that front, things are looking a little more lively in 2010 than at any recent time, so there's a lot to look forward to.
I realise this blog post got a bit long and rambling, so for that I apologise. I might follow up with some further thoughts later in the week. Thanks for reading this far, I'm off to the parents' place for a weekly roast dinner. Some things never change. :)
posted by Paul Montgomery | 1:16 pm
1 Comments:
We can never underestimate the financial benefits that the Tombstone icon has delivered.
Post a Comment
<< Home