Oliver Thylmann’s Thoughts

The question will be gaining importance over the coming months (for the digerati) and years (for normal people ;)). With the other shoe having dropped, and now Facebook banning Google Access for Friend Connect, things are hotting up. The short version, Facebook says that they do not allow Google access because people will loose control over their personal data when they do that. This has been proven to be untrue, with the simple fact being that there is no Facebook API through which Google can tell Facebook to add a new option to people can change their sharing preferences for Friend Connect connected sites within Facebook. People can do that on the site they shared their info on or within Friend Connect, but not on Facebook, which is a Facebook problem though. Scoble adds a few more things which are relevant though, coming back to the age old problem of syncing your data across several services, something that needs to be settled once and for all and probably never will. The problem is that Facebook descided for me that it is a problem for me, which I might not agree with. (Update: Later somebody from Google comments that even that is wrong. )

The problem comes down to the fact that Facebook does not want to loose control. Share your friends with Google Friend Connect and Google knows your friend and you can theoretically leave Facebook and take your friends with you. It’s not quiet there and but that’s what this might end up in. Michael has a great post about Data Portability on Techcrunch entitled “Data Portability: It’s The New Walled Garden” and he is very right. Let me quote one bit from Michael’s article:

Let me put this another way. How dare Facebook tell ME that I cannot give Google access to this data!

I am starting to wonder if there really isn’t a huge opportunity here, but still in the wondering phase.

There is one thing I would like to add though. I personally entrust my social network to Xing for example, and I will most likely entrust it to Google and there is one simple reason for it. Motivation on the side of the service I am entrusting it to.

Xing has a clear revenue model. I pay them a monthly fee to get a few extra services and with that model they get a wonderful EBIT margin. They are doing fine and I sure hope that Lars keeps his head on straight and doesn’t do anything irrational that removes the clarity from that mission, but I am pretty sure he will do the right thing. Xing’s motivation is in getting my data, getting my connections, having my friends on there, and getting me to sign up for more services. Great. Making it more freely available is something that Xing will have to deal with because when I can access it from anywhere, I will be happy to put everyone on there, also my non-member contacts while I am at it, making it my address book, but I wrote about that before. Huge opportunities.

For Google, they are an advertising company and they want information. But beyond that, they really do want all information to be freely available because that means that they will be able to run ads again that information, find it, give people better answers, … you know what I mean. that means that my social network is just a means to an end, not the end in itself for them, as the monetization sits somewhere else. That’s the beauty of Friend Connect. They help more social networks blossom and more content being created and content will win. Freeing up my data is in their best interest, or rather making it available.

For Facebook, there is no business model. I am not sure if making the data available to me is something that they will do because the only thing they have is that data. They have no business relationship with me, or do not have some other use for that data other than mining it and at some time coming up with an idea.

So who would you entrust your data with? Nobody (aka something like Noserub) or Google, or Xing, or … .

Things are really hotting up now in the social networking space and it goes in the right direction due to peer pressure. For the news I am talking about you can read this Techcrunch piece, but in short:

• MySpace launched Data Availability (see Mashable)
• Facebook answered with Facebook Connect
• Google is coming with something they call “Friend Connect” on Monday as of Techcrunch

While this is not real Data Portability it is a good first step that has actually been a lot more likely. The race is now going full throttle in terms of who will be your identity provider on the internet, meaning everywhere in the long run.

The basic idea is that you can use your MySpace/Facebook/Google information and give access to it on other sites. So I might register somewhere else and give it access to granular bits of my identity, giving the new service information about my friends, hobbies, most clicked ads, brand affinity, … you name it. This will go a lot further than simply moving your friends around. Especially because this is only Data Availability (with all 3) not portability, so you are not moving anything around.

The problem is that there will be one main provider most likely. If I attach all my new services to my Google account, Google will know about all those services and Facebook will only know about itself. Hence being at the center is important. Portability is not needed to convince the average Joe, but Availability helps making your the center.

That also means that building a social network gets easier as such, you just need to bring a unique value, bring something to the table that is not in the data but rather in the service you provide with that data. It is really nice seeing it all come together because it validates the choices we made at Ormigo in terms of development and furthers my belief in the long-term vision.

We have Hotmail, AOL Mail and Yahoo! Mail as players in the email space in the US (GMail is really still tiny among normal people ;)). So who will be the winners in Social Networking or Identity Data Provisioning. Facebook, or is it just a fun place to hang out at? The mail providers as they already know? LinkedIn, but they don’t have the mainstream. Plaxo might be interesting. Somehow I feel Google has the data to do it efficiently. In Germany we have United Internet with Web.de/GMX/… in email, so there might be another winner here, but will it be StudiVZ or Xing? Xing is still too small, StudiVZ not trustworthy enough (yet?). The race is still on.

I wrote about the ad market needing a change in July of last year, and now Emre Sokullu explained the idea very well in his post entitled Plan B for Microsoft: Split up the Advertising Atom.

He take the Microhoo deal as a starting point, thinking about how Microsoft can beat Google, but it’s not really something that is unique to Microhoo. The suggestion he has is something that is bigger than Google, bigger than any advertising system out there. Think VISA for advertising.

His idea is best explained via the picture below from his post:

The idea is that you split everything up, and create it around a standard so that the different silos can interact. The cool thing is that this is how our AdServer at Ormigo is structured, first because it makes for an amazingly efficient ad server that runs on Amazon’s Web Services Infrastructure, and it provides for some interesting new possibilities.

Our products that we advertise are within different objects stored on S3 (inventory silo) and the Placement Silo is really our AdServer that holds all the information about the Placements. The very cool thing about the above model is that you can have different inventories fill the same placement, or add on top of that different parameter silos that take care of optimization. Thing content match or behavioural targeting.

The cool thing is that you would enable Open Innovation. Of course there are lots of things still to be thought out to make it really open, meaning you make it a VISA model in which everybody can use everything and can do what they want with it but they will need to adhere to a set of rules and standards. This might make prices more transparent, and open up for real competition, which will not be good for everyone. But the thing is that it enables is a really open market place where different people can write a Placement Silo for TV ads that suddenly grab the right Ads from the Inventory Silo, optimized through a Facebook Parameter Silo to only show ads my Friends like. Who gets what? Who pays what? What about the wining and dining part of advertising? Lots of things not settled, but worth thinking about.

Lots of social networking stuff happening in the last few days. First Christos Cotsakos, former CEO of E*Trade, has raised $29.6 million for Moli, a new social network, after having put in something like $20 million himself. The company is already at over 50 people plus more offshore. GigaOm has a good review online, with the biggest special thing of Moli being the option to have several profiles folded into one, meaning you have one for work and one for family, with different information shared within them. At the same time, Techcrunch posts that Lookery is offereing advertising runs on Facebook Apps for $0.125 CPM, which is awfully low. Then chatty Mark Zuckerberg talks about numbers at a company  event, which leak: 2007 saw $150 million in revenue, 2008 will see $300-$350 of revenue with $50 million EBITDA but $200 million of CapEx (read buying servers and the like), so a loss of $50 million. If we take EBITDA like Earnings (just to make it easier because they will never buy stuff again … yeah right … never mind … let’s go on) then the $15 billion valuation would add up to a P/E ratio of 300! Forward P/E that is. Now that’s a valuation to look for!

The problem is that in the Google Earnings Conference Call, it just came out that monetization of social networks is really hard, advertising wise. Xing is in a very good spot here because they are monetizing via their users. Within our own platform, the social network is just a side effect that gets used. Pure play social networks will either have to become really really huge, or will most likely face a problem. At $12 cents cpm, even thinking a site is fully sold out, and gets 2 ads per page, that would result in $240 for a million site impressions, meaning $2.4 million of yearly revenues for a site like StudiVZ+SchuelerVZ. That is really nice, no questions asked, and a lot of nice things can come of it. The problem again might be that there is a social contract between the networks and it’s users so any new push to monetization in a new way might lead to a revolt of the users.

Facebook is surely on a good way, but it will be interesting to see their next investment round, in light of possibly further bad news, or an IPO to see what they are really worth. Microsoft couldn’t care less because the investment was more of an ad buy, but for the other investors, I don’t know.

It’s not all art, there is a lot of science to it, or so it seems. Check out this very well commented post on Techcrunch called The Secret Strategies Behind Many Viral Videos. So you think all of the top videos are there by accident? If you take a look at the most viewed videos all time there are a lot of brand videos on there for my taste. This made me wonder a few days ago how this happens and the article seems to be a good pointer.

In general, it’s simple. You spam, a quasi legal way. You have several youtube accounts and comment around, you post on forums, pay bloggers to post a video, use 5000-friends-accounts on facebook to distribute the video, … stuff like that. If the video then has a few qualities that the post talks about, you might have yourself a viral video. Congratulations