Messages Exchanged by POF users have grown 67% since Jan 1st to 100 Million last week. Pageviews are around 2 Billion per week, putting us over 100 Billion Pageviews a Year.
This graph only includes messages exchanged via the site and excludes chat messages.
Scaling is not very fun these days.
We have previously been sued in the US for using decimal numbers and another time for Matching people… It seems that we are now getting sued because someone somewhere on the internet ran an Advertisement for true.com using an image they did not have rights to. Our only involvement in this case happens to be that one of the places this marketer ran their ads was on POF.com.
I am pretty shocked this lawsuit was even filed -does this mean every Adsense publisher, newspaper or website can now get sued for the contents of the ads that appear on their websites?
After reading yet another story about how online dating can not predict who will have a romantic relationship I thought I’d say something. Its clear from reading the paper and these posts that Eli and his team do not know what POF is actually doing when it comes to matching. What we are doing is completely different than everyone else.
1. Similarity and other Psychological constructs are not used in our matching system. This is something a Psychologicist would do if they were responsible for making a matching system. In my opinion Psychological approaches to matchmaking are next to useless, they offer relatively small improvements over chance from all the testing we have done. We’ve hired the best Psychologicist’s to try and create a matching system and than validate it using our couples data. In the end Psychological evaluations and matching turned out to be completely useless. So we agree with Eli and team there.
Back in 2007 I went on the today show and revealed that I created a behavioral matchmaking system that helped match people up. Almost a year later Match.com hired a bunch of people to try and copy what i was doing and that seems to be their system today. This is a great system for getting people to send messages and in match’s case optimize revenues but it is a poor system for predicting relationships.
Given that behavioral matchmaking and recommendation systems were not powerful enough I started getting people who left the site to tell me who they entered into a relationship with. Using MILLIONS of people who left the site in a relationship we could very quickly determine what kinds of people got together in a relationship and with who. After about 6 months 45% of the relationships split up giving us a very clear picture of what kinds of traits work in relationships and which ones don’t. Keep in mind we make people answer hundreds of personality questions and we have all the data they entered on their profiles along with other stuff we collect. Using simple predictive models we are able to tell who you are most likely going to start dating with a very high degree of accuracy. If you do enter into a relationship with someone and it is a unstable relationship we could probably tell you the week in which your relationship will end.
Given what i’ve stated above there is no way that Eli and his team can argue that ”a mathematical formula can not identify pairs of singles who are especially likely to have a successful romantic relationship” Predictive modeling works in every other industry imaginable Online dating is no different. In short sites Like Eharmony, PerfectMatch and Chemistry are designed by Psychologists who think the world should work in a certain way and they build a matching system that enforces that worldview. At POF we just find successful couples and lets them be the basis for the matching system.
The following graphs the probability of a relationship breaking up. In order to be better than chance at predicting long term relationships all we have to is make sure we do not generate matches that lead break ups. Last year we could predict which relationships will fail with a greater than 80% accuracy. The relationships shown here are not the result of a matching system but relationships that randomly happened on POF. The sample size is around 21,000 and skewed to heavily to under 35. The way to read the graph is 42% of relationships formed on the week of 2/20/2011 are still in tact on Feb 13th 2012. That is to say neither the male or female came back to the site to signup for a new account.
In the last year POF has gone from having no mobile apps to 3 Billion pageviews a month on apps & 300 million visits and our mobile traffic continues to grow at 3% a week. In comparison the website has been around since 2003 and only has 160 Million visits a month and 4 billion pageviews. To give a sense of how quickly mobile is taking over online dating 40% of our signups are now via a mobile device in the US up from under 10% at the start of last year. Ipad usage grew 88% in one month from December 2011 to Jan 2012.
Now its great to have all this traffic, the only problem is now one has figured out how to make similar levels of money on mobile as the web, unless you do some real scammy stuff. So ya its great to have more traffic on mobile than every other dating app combined in english speaking countries but it doesn’t matter much if you can’t really monetize it at high levels and it starts to canabilize your web traffic. Given our current growth POF is probably going to be 60-70% mobile by the end of the year. Now most people think that mobile is just for young people, that is simply not true, we are already majority mobile for 30-40 year olds in the US and in a few months 40-50 is going to be majority mobile in the US.
I managed to visit 25 countries in 2011, highlights would be seeing the last of the mountain gorilla’s in rwanda’s cloud forests, safari’s around africa, Easter Island, Machu Picchu, Bhoutan, Anchor Wat and of course Egypt, hard to beat being fed wine while touring the 3400 year old temple of luxor by moonlight followed by dinner in the temple complex.
While all this was going on, POF nearly doubled in size last year we hit a record of 50,000 signups on sunday! We also got a tech team and some administrators, We launched in a bunch of non English speaking countries, created mobile apps, created our own chat system and redesigned the site a few times! Just goes to show you can run an internet company from anywhere.
On the business side we’ve invested millions into other sites last year some with millions of users. We acquired a couple of companies and formed a few some new ones outside of the dating space. We’ve set aside about 10M to acquire other companies this year or to invest in things. What i’m looking for are dating companies in the mobile space, non english speaking countries or something that can help us lift our average revenue per user.. IE a paid site we can send a lot of traffic to and it converts.
We are also looking for companies that have at least 100k users, if you don’t have 100k users it means you don’t have a product anyone wants. We are looking to help other companies grow very fast, we aren’t looking to help companies design/build a product and then figure out how to get users for them. What we look for is companies that have already have traction and can turn into large companies.
In November plentyoffish got nearly 6 billion pageviews and served out nearly 32 billion images in addition to that our IM servers handle about 30 billion pageviews a month. All of this traffic is handled by our 11 webservers of which we could probably get rid of 5 and everything would still work fine. All of this may sound like it costs massive amounts of money, but all hosting/cdn costs combined are under $70k/month.
What are typical hosting costs for other dating sites?
A lot of people have been setting up white label dating sites the last few years. The most important factor in making money doing this is something most new people don’t even realize they can change.
1. Never ever agree to the default terms a white label dating service offers you. By default the rates are as low as 50-50% rev share. We won’t even talk to a company unless the rates are closer to 80%-20. The majority of the major sites on these platforms are closer to that rate. 30% may not seem like much but if your site was earning 100,000 a year you now get paid $80,000 a year instead of $50,000 a year. If you had $45,000 in expensives that left you with $5,000 in profit. If you where at 80% revshare your profit is now $35,000 because you earned a extra $30,000
2. Constantly negotiate. Best example i can give is we were making $20/paid signup via adnetworks. We went direct and because of volume we were able to get paid $80/signup for the exact same users. When it comes to white label dating site the same rules apply, you can always increase revenue sharing in your favor. If you don’t you wont’ be in the space for very long as competitors are going to dramatically increase their revenue share over time.
3. Users are rebilled on average of 2 or 3 times. Some sites have different rates for new users verse rebills, for example you get a real high percentage for the first charge but a real low rate for the rebill. Lets say a user is billed 3 times for $100. If you make 80% on first sale and 50% for rebills that means you make $80 + $50 +$50 +$50. Always make sure both your initial and rebill percent rates are really high.
http://www.inc.com/inc5000/profile/okcupidcom
2010 Revenue:$4.7 million
2007 Revenue:$380,912
Employees:18
Acquisition price was 50 Million cash and 40 Million earn out. I’m surprised okcupid was able to make so much money given their 400M pageviews/month. They have 3 ad units per page and auto rebilling, compared to our 1 ad unit and no outrebilling on memberships.. Currently POF has around 5 Billion pageviews a month, if i monetize as much as okcupid that would be around 55-65M in revenues.
I was reading this article today about match talk about how they sent 1.2 billion emails in the last 6 years. It reminded me how far Plentyoffish has come, and how I’ve been doing all the database work except billing since the beginning. In a few weeks i’m going to give up much of that responsible and hand it off.
I can’t believe the scale of what’s been done. Its weird because day by day the growth isn’t that much but when you look back over the years its huge. Plentyoffish users send 5.5 Billion Messages a year.. That is 27 times what match.com users send. The hard part about scaling is making it cost effective. Match doesn’t have a problem as they earn 400M a year which works out to $2.00 an email sent on their system per year. For match buying new hardware is nothing. For plentyoffish we have 27 times the messages even paying a penny per email is not viable. But what truly takes computing power is our Behavioral Matching Engine, talked about it a bit in 2007 on the today show years before any other dating site even knew that kind of stuff was possible. I feed it with nearly 20 billion pieces of data that are stored in the database, and all of this is done in real time. That is what requires hardware and lots of it.
At the end of the day the site is so large, I need to give up the day to day running of the SQL databases. Not only do I not have as much time to pay attention to it but I also don’t know every aspect of SQL Server anymore. Its funny one day you knew everything there was to know about SQL server because you had to work with it daily but than a year goes by, or 2 a few upgrades and suddenly you only kind of know whats going on.
According to IAC….
54 million singles in US
2.6 million paid at any time.
5.5 free and non paying.
8.8 million singles have used online dating.
http://ir.iac.com/eventDetail.cfm?EventID=97073
Only 5%of the population is using a paid dating site at any given point in time and another 5% a free dating site. The real focus of most of the top dating sites is how to get the other 90% of the population to start using dating sites.
