Archive for February, 2008
Ask the readers: If you could drop 100M users onto any site, what would you build?
A great hypothetical question…
My cofounder Matt Rubens recently posed a great question to me over dinner:
Let’s say you had unlimited access to internet traffic, and could literally dump 100 million unique visitors in a single shot onto a site – what would you build?
I honestly don’t know the answer to this question, but here are a couple considerations:
- Any vertical site, be it pets or cars or even gaming, would ultimately lead to a huge % of the 100 million immediately bouncing and being lost – this might be OK, if the category is high value enough?
- Conversely, communication is probably the most horizontal application you could do, but it’s low monetization and not clear that you can go build in more depth after people think of you as a communication site
What do you guys think?
I would love any ideas or insights you guys have on this question!
Please comment to this post or send me an email at voodoo [at] gmail.
Reader question: Why answer questions?
A: There are probably 3 main drivers:
First, I like being helpful – I know there are problems that folks are facing which, if they had a short conversation, could help them skip months of turmoil. For some very specific topics, like building an ad network, or being an Entrepreneur-in-Residence, I can help with that. And I enjoy it!
Second, writing this blog is very much about honing my own thinking. So when folks ask questions that are challenging or expansive, it gives me an excuse to exercise my brain to come up with a reasonable response.
Third, when I get specific questions from people, it removes the abstraction of publishing to 2500 subscribers where I see the number, but don’t know anyone’s names or have any additional profile information.Â
Reader question: About college…
A: A lot of super smart people I went to college with had a very good idea of what they wanted to do right away, started working towards it, and are still on that path 10 years later. (Most of my Early Entrance Program classmates are now in med school, grad school, etc.)
… but not me :)
Instead, it took a while for me to learn that it was okay to go off the beaten path, and that I should go chase whatever I’m passionate about. However, even in Seattle, there aren’t many role models who can teach you the right way to manage that process. I would ultimately receive a lot of help from two Bills – Bill Ericson and Bill Gossman – who did teach me a lot about this.
In short, I wish that I would have had a more “strategic” view of how college fit into my life, as a tool rathe than as a monolithic step that I had to get past to get anything done.
(That said, stay in school kids!)Â
Reader Question: Are there new opportunities in social networking?
A: First, let’s talk definitions
I get variations of this question a lot, both online and offline. And everytime I’m asked, I tend to ask several followup questions, like:
- What do you mean by “social network”?
- What do you mean by “make it big?”
Let’s talk about the first issue, the meaning of a social network. One could define this as any site that has a standard set of pages like a profile page, a friend list, a newsfeed, a wall, and all the other mechanics. Another definition might be to do it more narrowly, like a set of mechanics and also the primary use case of communication. Yet another would be to define it as a site supportin the OpenSocial API.
It’s obvious, but the more narrowly you define the term, the harder it will be for a specific site to replicate its success. If you completely cloned Facebook today and released it to the wild, it would have a vastly different outcome than doing it 5 years ago.
Similarly, the concept of “making it big” is nebulous as well. For many entrepreneurs, making a company that throws off $5 million a year that supports a great lifestyle is an incredible outcome. Many indie Facebook developers and associated “lifestyle entrepreneurs” are primarily driven by that. For other entrepreneurs, they aspire for a billion dollar outcome.
My take on credible social network businesses
With all that said, I do think there are many opportunities for people to build sites with social networking backbones, and attract 10s of millions of users.
The reason is that there are simply too many people in the world interested in mega-niche topics, like:
- Fashion
- Knitting
- Board games
- NASCAR
- Kittens
- Photography
- Comic books
- Import cars
- etc.
And the same way that there are specific ecosystems of clubs, stores, student organizations, magazines, etc. that specifically cater to these niches, there will be websites that cater to these and will find some way of reaching millions of users.
I think where you run into problems as a business is when you are focused on a site that:
- Is contextless and primarily is used for communication
- Is reflective of your actual friend group (not new people)
- Is people-centric rather than centered on media, games, etc
Taking all of this out makes it so that your site not only supports the same use cases as Facebook and MySpace, but also has the same “feel.”
There are many permutions of this that can work well as a social network – it’s worth exploring more.Â
Trying out a new widget called Askablogr
For people who have this on RSS, please click to see my blog…
I’m trying out a new widget called Askablogr, which appears on the left pane. If you click on it, you can ask me a question, which I’ll then answer. (Maybe)
Try it out and let me know what you think!
I’m sick :( Also, going to Game Developers Conference for the week
Sorry for the slow blogging – I’ve been sick the last couple days.
This week, I’ll be at the Game Developers Conference held here in San Francisco. Couple predictions:
- The place will be SWARMING with VCs – there were only a few last year, but this year it’ll be over the top due to funding in the last 9-12 months
- Mobile games will continue to be dour, and the folks there will be jealous of all the new distribution opportunities afforded by Facebook/MySpace/etc.
- Internet people – that is, folks with backgrounds in Rails/SEO/SEM/widgets/etc – will in also be invading. These guys have never really built deep games before, but they sense an opportunity to build some simple content and use some new fangled techniques to distribute them
- "Club Penguin for X" has replaced "YouTube for X" as the new hotness
Also, here are a couple bloggers who write about games that you should read, if you’re not:
- Erik Bethke – GoPets
- Dan Cook – Microsoft
- Jeremy Liew – Lightspeed Venture Partners
- Daniel James – Three Rings
- Charles Hudson – Gaia
- Sean Ryan – Meez
- Nabeel Hyatt – Conduit Labs
I will be switching between the casual games summit and the worlds in motion summit in the first couple days, followed by attending the typical fanboy presentations on my favorite games. Should be fun!
From Seattle? Please introduce yourself…
I’m visiting Seattle later this month, week of Feb 25th
After almost a full year of not visiting the ol’ hometown, I’m finally visiting the area to see family and friends.
If you’re a reader of this blog, or if you have recommendations for folks I have to meet, please shoot me a note at voodoo[at]gmail, and include:
- Couple sentences about yourself and your background
- What are you working on?
I’m particularly interested in meeting entrepreneurial engineers who are building cool stuff.
Thanks!
Nasty ad on Facebook
A friend of mine sent this ad to me – on the top, you’ll see an ad that says:
, You Have 2 New Unread Messages
One Of Your Friends Has A Secret Crush On You. Find Out Who!
Combine this with the "CONTINUE" button next to it which matches the Facebook interface, and you get something pretty nasty. I assume that in the case where the ad network passes in your first name, they are able to personalize this message.
As I’ve said in the past, ad people come up with some very creative stuff – I still consider the geo-targeted Adult Friend Finder ads to be some of the most technologically advanced stuff online right now.
Diminishing returns on social media virality?
Question for the audience:
Response rates are falling on notifications, newsfeed items, etc. There’s a point at which it comes very very hard to be viral on top of Facebook anymore – this is starting to be true with e-mail virality, for example.
This roadblock will happen one day, but how long before we hit it? Months? One year? Years?
Folks with better visibility into Facebook viral stats would know better than I do…
BTW, I think this is a great reason for very app developer to be looking at OpenSocial right now, because it’s all green fields, and people who invest the time on things like language, cultural understanding, etc., will have a huge advantage in being able to generate the highest growth apps.
What’s happening with the top Facebook apps?
I collected the charts for the top Facebook apps and posted them below – overall, there seems to be a general decline. Note that this may NOT be a bad thing – perhaps developers are getting incented to create deeper and richer apps – more details below.
Anyway, looking at these charts: Why is this happening?
A couple thoughts:
- More apps means more competition (even for the top apps), which creates fragmentation – it become harder to keep any good thing going
- Changes to the Facebook viral channels are continually driving down the response rates, incenting developers to create stickier apps that can retain more active users
- Continual use/abuse of the Facebook viral channels are making users cynical, driving down response rates
- The top apps have hit the "carrying capacity" of the Facebook userbase, and have saturated the entire network – thus, the only option is for the usage to decay over time, since more viral growth isn’t an option
There’s probably multiple explanations here – note that for some of this, Facebook may be "guiding" the developers to create more meaningful applications via this "active users" metric, so even if the top apps are losing users, that’s OK.
Question for FB app developers: How do you avoid becoming a 3rd-tier Slide.com?
Strategic question for Facebook developers
During lunch today, an interesting question came up amongst a bunch of Facebook developers:
How do you avoid becoming a 3rd-tier Slide.com?
It seems like beyond the first couple players – Slide, RockYou, and others, there’s a huge drop-off for small teams who have raised <$1MM and trying to compete. Problem is, when you look at a company like Slide, they have 10-25X the number of people, 100X the funding, 100X the valuation, and 1000X+ the reach.
I’d be curious to hear how people are thinking about this?
My main thinking about this would be to pick some type of vertical – either by industry, demographic, media type, or something else that raises the bar for defensibility. It makes it much harder to be a "fast-follower." Anyone who is working on a purely horizontal communication basis will run straight into the Slide/Rockyou wall.
Any other ideas here?
The design of social spaces
Designing social spaces
Earlier today, I had a great conversation with Rashmi from Slideshare, where we dug into the design of social spaces – inspired by that, I wanted to write down a couple ideas that came up over coffee. Personally, I draw a lot of inspiration from offline social spaces, particularly malls, bars, cafes, stadiums, etc. There are places where these analogies hold, and places where they break down completely.
Here are some of the interesting design decisions for social spaces:
- What is the "unit" of your social space? (people/discussion/media/etc.)
- Is your social space open, closed, and somewhere in-between?
- How personal versus mainstream is the content on the site?
Let me emphasize that there are advantages and disadvantages to all of these different approaches – I’m sure there are many companies that are built using every permutation of this.
Anyway, to drill into each question in more detail…
What is the "unit" of your social space?
Every social site has a fundamental unit that’s tied together through social context. For MySpace and Facebook, that’s clearly people. For YouTube, Metacafe, and others, it’s video. For forums, it’s discussion-centric.
The easiest rule of thumb for understanding which unit dominates your social network: What is the core action your users are repeating over and over? If it’s navigating from video to video, that’s one thing, but navigating from person-to-person indicates another. For many sites, even when you are looking at a person’s profile, the most emphasized navigation element is to draw them back into a piece of media – that indicates that it’s media-centric rather than people centric.
An interesting example of this is Facebook photos – when you click into a person’s profile, you can then click through a whole bunch of photos. But ultimately a person "contains" a set of photos, and it’s not a "YouTube for photos" situation where I might be looking at a photo from Person A’s profile, and it lets me directly navigate to related photos from Person B’s profile. Instead, I always have to go up a layer from photos to the containing person, and then go to another person, then go down into more pictures. This emphasizes Facebook’s people-centric approach.
In my humble opinion, people-centric social networks tend towards communication (or pseudo-communication, like poking), which in turns tends to drive more stickiness. The reason is that when someone sends you a directed piece of communication, that’s the most urgent possible way to get you back to a site. Compare this to a piece of media that may or may not be directed to you – you might find it compelling, yet it doesn’t drive a huge sense of urgency.
Is your social space open, closed, and somewhere in-between?
Another key design decision has to do with whether or not your social space is open, closed, or somewhere in-between. If it’s open, then everyone can browse everyone else’s profile (or at least folks are visible). If it’s closed, you have to be invited to even see a single person, and vice-versa.
And there are examples like Facebook or Yahoo Groups which tend to be semi-open (or semi-closed, depending on how you look at water in cups). In those, you are able to browse people in your "network," but you can’t browse further than that.
By far, the most interesting issue here has to do with the process in which people self-identify with a web property or not. If it’s very open, then it makes it easy to say "oh these aren’t my people." The advantage of open, of course, is that you get some network effects as people join – you can reach a critical mass more easily (even if it’s in Turkey). On the flip side, if you are too closed, it can be harder to get started, but your site will seem more accommodating for a larger span of demographics.
How personal versus mainstream is the content on the site?
Photobucket, Flickr, and other photo sites ultimately hold a very different kind of content than YouTube. The reason is that the former has mostly "personal" content which is appealing only if you know the folks involved – sometimes this is combined with the term "long tail UGC." Compare this to YouTube, which has more mainstream content in the form of pirated media, mainstream studios releasing content, etc.
Of course, if you have personal content, it’s then important to either have a social graph of your own, or to sit on top of someone else’s. Otherwise, if you don’t provide an easy way for your friends to see your photos and vice versa, then the personal content is just sitting there. Of course, if you have both components – both personal content and the personal social graph, then you have a very sticky site.
Mainstream content is great, but lacks personal relevance. It doesn’t need a pre-existing social network to exist, yet it’s less relevant and not very urgent for users to consume the media. Of course in the case of YouTube, they exist on so many social graphs (MySpace/FB/Email/IM/WOM/etc.) that it doesn’t matter.
Couple attempts to classify successful consumer sites?
Here are a couple versions I’m thinking:
- YouTube: video-centric, open, mainstream+personal content
- MySpace: people-centric, open, mostly personal content (except for hot girl,s which is mainstream!)
- Facebook: people-centric, semi-closed, personal content
- SomethingAwful: discussion-centric, semi-closed, mainstream+personal content
- ICanHazCheezburger: cat-photo-centric, semi-closed, mainstream cat content :)
Am I missing anything from my framework?? I suspect there are a couple more axes I am leaving out.
Top keywords for how people find my blog
I always find the sources of traffic for my blog interesting – I thought I’d publish them here.
If you’re a blogger, you should go to your Google Analytics and hit Traffic Sources, then Keywords, and do an Export to Excel. Would love to see yours – please post it to your blog!
Couple observations about my list:
- Interesting that a lot of people are searching for me by name! Weird, although obviously there are probably tens or hundreds of thousands of "Andrew Chen" people out there, now and historically. Just at my college there were 5
- As to be expected, there’s a ton of folks who are searching about "Entrepreneur-in-residence" from my most recent gig
- Similarly, there’s a lot of folks who are interesting in advertising, viral marketing, and the typical digital media stuff I write about – hopefully they enjoy the content they see
- Also, interesting enough, there are a bunch of consumers who are looking for Diner Dash, or Facebook, or MS Paint, which are all consumer products I’ve mentioned in the past – those guys are probably very disappointed at what they find on this blog!
Here’s the first couple dozen of my list for 2007 to now:
- andrew chen
- entrepreneur in residence
- entrepreneur-in-residence
- andrew chen blog
- entrepreneur in residence""
- vertical ad networks
- viral loop
- futuristic play
- vertical ad network
- andrew chen""
- facebook dating
- andrewchen widget
- andrewchen reward schedule
- new diner games
- how to draw on ms paint
- sketchfu
- how to draw with ms paint
- andrewchen
- social monetization
- how to draw in ms paint
- club penguin viral
- game designer tutorial
- ideo method cards
- how to create a social network
- new diner dash
- first impressions ""andrewchen""
- google ecosystem
- futuristic
- ironic article
- widgets = ad networks
- andy chen
- game design tutorial
- uploading andrewchen
- 10 ways to ruthlessly acquire users
- 1 9 90
- 1-9-90 rule
- andrew chen mdv
- facebook monetization
- andrew chen viral loop
- facebook and dating
- site:andrewchen.typepad.com viral