Category Archives: socialgraph

Noise Aggregators

bored
Jevon has a great post up today on how the new trend towards social aggregators is missing the point.

I’ve called these things underpants business models. Collecting a lot of people’s seemingly-useless personal stuff in one place, but then, through some magical step 2, everyone profits. I, like Jevon, and everyone else, are still waiting for anyone to come anywhere close to step 2. Yet the aggregator startup fad just keeps coming. Everyone seems to be wearing them these days… Jaiku, plaxo, friendfeed, socialthing etc.

But, as Google has proved since 1998, there’s little value in aggregating the whole internet, but a lot of value to be created in synthesizing it. Aggregating random data easy. Relevant, human-interesting synthesis hard, but the only place where value is created.

What exacerbates the synthesis problem is that, often with these apps, even the underpants suck. Most of these sites aggregate feeds based on any/all of what feeds your friends have available rather than what’s meaningful. No thanks, I don’t need an unfiltered list of every single track you listened to in iTunes yesterday.

I think only facebook -in it’s heyday, before the application spam took over, and drowned out all semblance of meaningful social interaction- ever had it close to right. (The original) facebook filtered heavily and focused only a few relatively high-social-value feeds of information. Even this has been debated.

As Bruce Sterling put it a year ago following a lot of blogs/twitter/social-presence can be like watching someone getting beaten to death with croutons.

Or to keep with the cruder underpants metaphor: Nobody wants an experience that feels like drinking from the firehose social diarrhea.

Anyway, thanks to Jevon for pushing me towards getting down the real post I’ve meaning to write…

Coming up on ThomasPurves.com : Why the Social media is still absolutely important, and why it don’t matter at all. (a.k.a. How socialness is just one of many vital prerequisites, but not the point, of the future of software)

What you need to know about FaceBook

fbcamp crowd
My home town Toronto just held it’s second ever FacebookCamp to another near-soldout crowd. I’d say approaching 250. You really can’t swing a cat in this town without someone tagging the cat on facebook.

Ami Vora form facebook led off, followed by a rather exhaustive series of talks and demos from local facebook app builders and marketers. of note:

At Harvard., Amy Vora maintains 43 friends

This is a just small fraction of 43M active users this week. 3% growth each week
1 in 5 canadians on facebook

Canada is wildly overrepresented on facebook and will be soon a good case study of what happens after you saturate a country (will the book still be as interesting without a constant influx of new faces?).

Canada is bigger than UK. 20% of all Canadians are on facebook (something to do with long winters?).

International and non-english countries are not on the radar. They say internationalization is coming soon though. Yes, some people in the world don’t speak English first. The bastards.
Demographics are different everywhere except the US. US or A is totally college dominated, everywhere else is not. There was a pretty chart.

FB is creating the social graph of how ppl interact (really?). Everyone thinks social graph is a stupid name.

Social graph powers massive more usage of facebook photos/events than flickr, evite
From a scientifically rigorous 3-person sample size of a facebook HQ lunchroom snapshot, I can reveal that everyone that works at facebook is also under the age of 23. This may explain a lot.
If the social graph is so important why are the relationship descriptors so weak? (We hooked up once, it was awesome has been pointed out as a relationship descriptor you may/may-not want to use amongst your professional colleagues. Or after the age of 23. Depending I guess on exactly what your profession is. The word from Amy is that facebook might be working on this.

As a platform Facebooks thinks about providing
1) Gowth (powerful viral distribution)
2) Engagment (all those great feeds and alerts)
3) No business model. Sorry kid, you’re on your own to figure that out.

Facebook is also a little confused about this idea of monetizing apps. They list their granting fund and VC as options. No, these are financing options not business models.

Advice on how to provide engaging content
• Relevant info
• Showcase interaction
• Usability
• Fresh content
• Use integration points well
• Develop and iterate, low investment, quick feedback, lots of control
• Think of intelligent promotion
• Incorporate privacy

And this is just the beginning. Only 4 months in to platfomification of facebook

Here’s how to make money on facebook

How to monetize by Greg Thomson of My Garden and My Aquarium fame:

You need at least 250k installs, 10k users to start making money

$3.00 an active user is worth per year [really? To be safe assume 20% of your users are worth $3]

He’s at 8M installed users

Start with a base of ad-support: Appsaholic pay per click 0.10/click. Adsense 0.07 click

CPM rates of about $2. Depending on page view frequency per user

Most of the popular apps are entertainment

Ads peak after a day or two. Keep them fresh. Constantly need to keep monetization and ads fresh.

A/B testing has huge benefits. [flip side: This isn’t a turn it on and walk away money making machine]

Selling your app. Never happened. $1 install, $8-$10 per active user

Have a base of banners and sell incentiviszable offers. they can earn “coins” to spend on fish for their tank if they click on surveys for which you earn commission. (see your favourite affiliate ad network)

What’s it take to support 8M users in this case? 2 servers. 10 rendering servers uploading flash to jpgs on s3

Don’t spend more than a week(!) on an app, or the risk goes up

How experienced this grizzled veteran of the blue and white book? About 3 months now.