Everyone,
I apologize in advance for the length of this message as it is going to cover a lot of topics.
I have not posted anything substantive in a while, so I thought it might be helpful to touch upon a few issues that I think will be of general interest. I am going to touch on three topics:
a) who I am and who I am speaking on behalf of
b) some lessons I have learned in the last five years about offshore development
c) What we are building at Online.tv and how it might save you time and money
A: Who am I?
I think it will be useful for some of the newer folks for me to clarify who I am.
I am Antonis Polemitis. I run Ledra Capital. (www.ledracapital.com). We are a venture capital firm focused on higher education and online media and over the last three years we have built an very large position in .tv domains that we plan to develop into online .tv channels.
We put up Election.tv up last year as a test site and, at the time, were using National.tv as our overall .tv brand name. We were fortunate enough to acquire Online.tv last summer and that is the brand name we will be going to market with. National.tv will be retired.
I am sometimes asked: “Who is NetworkGlobal.tv; Oztralia.tv; Broadway.tv”. The Network Global sites are run by Gunnar Larson who was a former student of mine on one of our study abroad programs. Old-timers might remember him from our briefly running morning election news show on Election.tv called “Gunnar in the morning”. Gunnar caught the .tv bug and is running a few sites of his own, but there is no other formal affiliation between the sites.
B: Development – Lessons Learned
CPH: I am constantly being asked questions about development and in most cases I am being asked the wrong questions. The usual question is what is the Cost Per Hour?
The answer I give people is that it does not matter, but then nobody believes me and thinks I am trying to hide the answer. Let me explain further. I have used developers in Delhi, Bangalore, Nicosia and Bucharest. Cost per hour across those locations and across the different roles that people play (from junior developer to systems architect to quality assurance to design) ranges from about $15 per hour to $60 per hour depending on the firm and the role. The $15 is increasingly a rare sight at any reasonable firm in any location. Rates of $35 to $50/hr are now common. All else being equal, bigger firms charge more than smaller firms and Eastern European firms are more expensive than Indian firms.
So the first thing to note that even though this is “offshore”, this is not cheap labor. $40/hr is an annualized $80,000 per year. Furthermore, in the last five years I have been doing this, due to the falling dollar and the increased competition for labor, CPH has been headed in one direction only: UP.
I will submit to you, however, that CPH is the wrong metric. The number one factor that matters in keeping development costs reasonable is programmer productivity as a good programmer is several times more productive than a poor programmer and that is an almost impossible thing to judge in advance as there is great variance both between firms and even between programmers within firms. We work with dedicated teams so that over time we can build a complementary, productive team that knows our likes and dislikes and with whom I can communicate with effectively.
The Long Tail: One thing to keep in mind in terms of custom development is that it is easy to check-in and very hard to leave.
Both at Ledra and at my prior job, I had the pleasure of being involved in development projects that ranged from $50,000 and a few months of work to $50M and a few years of work (building a mortgage origination platform). They all shared one characteristic. The time and cost estimate that you start with generally bears no resemblance to what it will actually cost you to develop the project.
There are two issues that cause this that are not anyone’s fault – they are facts of life
a) It is humanly impossible to start with a blank piece of paper and design the exact application you want from scratch. As soon as version 1.0 is up, you are instantly going to realize that there are three more features you want, two more reports you forgot and hey, I don’t like how that flows. You can’t beat this.
b) While you were building your application, the competition was not standing still so user interfaces have become more modern, new features have been added, etc, etc, etc.
Personal Involvement: Your involvement in managing any development effort is critical. You need to expect to have to spend time every week reviewing specs, features, bug-testing if you want your application to turn out the way you want it. A long time ago, I used to be a fairly good C++ programmer. I couldn’t code my way out of a paper bag anymore but it is extremely helpful in being able to translate what I want into something that a developer can do something with.
WorkFlow – 90% of Iceberg: At my prior job, I conducted due diligence on a variety of specialty portals and was always surprised to be reminded of how time-intensive it is to run these. I know the study abroad field well, so look at www.studyabroad.com and www.goabroad.com which are each run by friends of mine. Each of these are part of network of a handful of websites and in total they employ dozens upon dozens of people to do things like invoicing, content management and customer support.
In other words, even in these “tech” businesses, business staff is the overwhelming majority of your costs. The same is going to be true in building out a .tv domain.
If you have not thought about these issues when thinking about development, then I submit to you that you are missing the boat on what is actually going to matter in your cost structure.
What does this mean? Building a nice-looking front-end template to show videos is a piece of cake. However, it doesn’t scale into operations at all. You have to bite the bullet and built the ERP / workflow behind it if you want to be able to operate your domain
Net Conclusion: Here is what I would use as a rule of thumb for development of a best-case individual site, done off-shore. If you don’t have a) time, b) a fairly good sense of what you want and b) > $50K+ over a year’s time to build out and add features, then custom development of a "real" web app is not for you. I see plenty of people spend $20K, run out of money/time/energy and have something that half works and is getting more obsolete by the day. Don't let that be you. You are MUCH better off parking and reselling than doing that.
If you are trying to do multiple sites, then you need a platform. If you need a platform, you better understand something or other about scalability and you should be expecting a minimum mid-six digit to seven digit investment before you are done.
C. What are we up to and how it might save you time and money
What we are up to: You will have noticed that we have been very quiet for the last few months. That is because we learned what we wanted to learn from our test sites and our conclusion was that stand-alone sites and even video publisher platforms like brightcove were not going to scale up the way we wanted from a workflow and system admin and new feature perspective across a portfolio of sites.
So we are building from scratch a high performance, scalable, set of web services (call it a “platform” if you like) that is going to allow us to build out a network of hundreds of online TV channels on our portfolio over the next 3-4 years. This is not an exercise for the faint of heart. It is the most technically sophisticated work I have ever managed and we are going to be working at the cutting edge of distributed computing to make this work and scale.
Our platform is going to cover: a front-end highly customizable design of the website, the back-end database, workflow and ERP, ad-serving and customized video streaming. It is going to be built on an open architecture so that it can be used, without any fear of lock-in, by parties outside of Online.tv.
Why might be it be relevant to you?
a) If you are a small to medium .tv domain holder, it almost certainly makes sense to use something like our platform to develop our your domains. I can basically guarantee you that the cost to you and time-to-market will be lower than development
b) If you are a large domain holder and are having “platform” thoughts of your own, I strongly encourage you to come have a chat with me. I am tapping into my network in the US to pick the brains of some of the best minds in scalability. At a minimum, I might be able to save you from some dead-ends; in a best case, I might be able to save you hundreds of thousands of dollars
Our goal is to take away “development” as a barrier to the .tv space exploding; both for our internal purposes (primarily) and for others as well (We feel that the more successful .tv websites that are out there, the better it is for us). I believe (and of course I might be delusional, but I doubt it) that this can only be done by a scale operator that can commit serious amounts of capital to this effort.
If you want to see something fun, do this: Our custom player is not fully built out yet, but our flash streaming is, so check it out here.
www.online.tv/demo
username: onlinetv
password: viewer
If you have a fast connection, watch the Ratatouille and Horton videos (they are the nicest quality) on full-screen mode. At least in New York, it is the nicest HD streaming I have ever seen .
We expect to have the video web services available in the end of May; the first version of the rest probably in the summer. Please feel free to PM me with questions or thoughts. I would love to have some of you as beta testers.
Thanks,
Antonis
I apologize in advance for the length of this message as it is going to cover a lot of topics.
I have not posted anything substantive in a while, so I thought it might be helpful to touch upon a few issues that I think will be of general interest. I am going to touch on three topics:
a) who I am and who I am speaking on behalf of
b) some lessons I have learned in the last five years about offshore development
c) What we are building at Online.tv and how it might save you time and money
A: Who am I?
I think it will be useful for some of the newer folks for me to clarify who I am.
I am Antonis Polemitis. I run Ledra Capital. (www.ledracapital.com). We are a venture capital firm focused on higher education and online media and over the last three years we have built an very large position in .tv domains that we plan to develop into online .tv channels.
We put up Election.tv up last year as a test site and, at the time, were using National.tv as our overall .tv brand name. We were fortunate enough to acquire Online.tv last summer and that is the brand name we will be going to market with. National.tv will be retired.
I am sometimes asked: “Who is NetworkGlobal.tv; Oztralia.tv; Broadway.tv”. The Network Global sites are run by Gunnar Larson who was a former student of mine on one of our study abroad programs. Old-timers might remember him from our briefly running morning election news show on Election.tv called “Gunnar in the morning”. Gunnar caught the .tv bug and is running a few sites of his own, but there is no other formal affiliation between the sites.
B: Development – Lessons Learned
CPH: I am constantly being asked questions about development and in most cases I am being asked the wrong questions. The usual question is what is the Cost Per Hour?
The answer I give people is that it does not matter, but then nobody believes me and thinks I am trying to hide the answer. Let me explain further. I have used developers in Delhi, Bangalore, Nicosia and Bucharest. Cost per hour across those locations and across the different roles that people play (from junior developer to systems architect to quality assurance to design) ranges from about $15 per hour to $60 per hour depending on the firm and the role. The $15 is increasingly a rare sight at any reasonable firm in any location. Rates of $35 to $50/hr are now common. All else being equal, bigger firms charge more than smaller firms and Eastern European firms are more expensive than Indian firms.
So the first thing to note that even though this is “offshore”, this is not cheap labor. $40/hr is an annualized $80,000 per year. Furthermore, in the last five years I have been doing this, due to the falling dollar and the increased competition for labor, CPH has been headed in one direction only: UP.
I will submit to you, however, that CPH is the wrong metric. The number one factor that matters in keeping development costs reasonable is programmer productivity as a good programmer is several times more productive than a poor programmer and that is an almost impossible thing to judge in advance as there is great variance both between firms and even between programmers within firms. We work with dedicated teams so that over time we can build a complementary, productive team that knows our likes and dislikes and with whom I can communicate with effectively.
The Long Tail: One thing to keep in mind in terms of custom development is that it is easy to check-in and very hard to leave.
Both at Ledra and at my prior job, I had the pleasure of being involved in development projects that ranged from $50,000 and a few months of work to $50M and a few years of work (building a mortgage origination platform). They all shared one characteristic. The time and cost estimate that you start with generally bears no resemblance to what it will actually cost you to develop the project.
There are two issues that cause this that are not anyone’s fault – they are facts of life
a) It is humanly impossible to start with a blank piece of paper and design the exact application you want from scratch. As soon as version 1.0 is up, you are instantly going to realize that there are three more features you want, two more reports you forgot and hey, I don’t like how that flows. You can’t beat this.
b) While you were building your application, the competition was not standing still so user interfaces have become more modern, new features have been added, etc, etc, etc.
Personal Involvement: Your involvement in managing any development effort is critical. You need to expect to have to spend time every week reviewing specs, features, bug-testing if you want your application to turn out the way you want it. A long time ago, I used to be a fairly good C++ programmer. I couldn’t code my way out of a paper bag anymore but it is extremely helpful in being able to translate what I want into something that a developer can do something with.
WorkFlow – 90% of Iceberg: At my prior job, I conducted due diligence on a variety of specialty portals and was always surprised to be reminded of how time-intensive it is to run these. I know the study abroad field well, so look at www.studyabroad.com and www.goabroad.com which are each run by friends of mine. Each of these are part of network of a handful of websites and in total they employ dozens upon dozens of people to do things like invoicing, content management and customer support.
In other words, even in these “tech” businesses, business staff is the overwhelming majority of your costs. The same is going to be true in building out a .tv domain.
If you have not thought about these issues when thinking about development, then I submit to you that you are missing the boat on what is actually going to matter in your cost structure.
What does this mean? Building a nice-looking front-end template to show videos is a piece of cake. However, it doesn’t scale into operations at all. You have to bite the bullet and built the ERP / workflow behind it if you want to be able to operate your domain
Net Conclusion: Here is what I would use as a rule of thumb for development of a best-case individual site, done off-shore. If you don’t have a) time, b) a fairly good sense of what you want and b) > $50K+ over a year’s time to build out and add features, then custom development of a "real" web app is not for you. I see plenty of people spend $20K, run out of money/time/energy and have something that half works and is getting more obsolete by the day. Don't let that be you. You are MUCH better off parking and reselling than doing that.
If you are trying to do multiple sites, then you need a platform. If you need a platform, you better understand something or other about scalability and you should be expecting a minimum mid-six digit to seven digit investment before you are done.
C. What are we up to and how it might save you time and money
What we are up to: You will have noticed that we have been very quiet for the last few months. That is because we learned what we wanted to learn from our test sites and our conclusion was that stand-alone sites and even video publisher platforms like brightcove were not going to scale up the way we wanted from a workflow and system admin and new feature perspective across a portfolio of sites.
So we are building from scratch a high performance, scalable, set of web services (call it a “platform” if you like) that is going to allow us to build out a network of hundreds of online TV channels on our portfolio over the next 3-4 years. This is not an exercise for the faint of heart. It is the most technically sophisticated work I have ever managed and we are going to be working at the cutting edge of distributed computing to make this work and scale.
Our platform is going to cover: a front-end highly customizable design of the website, the back-end database, workflow and ERP, ad-serving and customized video streaming. It is going to be built on an open architecture so that it can be used, without any fear of lock-in, by parties outside of Online.tv.
Why might be it be relevant to you?
a) If you are a small to medium .tv domain holder, it almost certainly makes sense to use something like our platform to develop our your domains. I can basically guarantee you that the cost to you and time-to-market will be lower than development
b) If you are a large domain holder and are having “platform” thoughts of your own, I strongly encourage you to come have a chat with me. I am tapping into my network in the US to pick the brains of some of the best minds in scalability. At a minimum, I might be able to save you from some dead-ends; in a best case, I might be able to save you hundreds of thousands of dollars
Our goal is to take away “development” as a barrier to the .tv space exploding; both for our internal purposes (primarily) and for others as well (We feel that the more successful .tv websites that are out there, the better it is for us). I believe (and of course I might be delusional, but I doubt it) that this can only be done by a scale operator that can commit serious amounts of capital to this effort.
If you want to see something fun, do this: Our custom player is not fully built out yet, but our flash streaming is, so check it out here.
www.online.tv/demo
username: onlinetv
password: viewer
If you have a fast connection, watch the Ratatouille and Horton videos (they are the nicest quality) on full-screen mode. At least in New York, it is the nicest HD streaming I have ever seen .
We expect to have the video web services available in the end of May; the first version of the rest probably in the summer. Please feel free to PM me with questions or thoughts. I would love to have some of you as beta testers.
Thanks,
Antonis
Last edited:
