startups, technology and fishing

I’m testing two new Feedjit widgets on the right side of this page. We launched the back-end code for these tonight and they’ll only appear on the public site in a week or so after they’ve been thoroughly tested.

The first is a simple page popularity widget that has been much-requested by our users. It shows the most popular pages for your website today.

The second is my favorite - it’s a passive collaborative filtering widget. It suggests other pages your visitors may like based on the current page they’re viewing and the traffic patterns of previous visitors. I’ll be adding more detail about the widget and algorithm in the FAQ once published.

If you’re an HTML hacker and an early adopter type, you can view/source of this page and grab them for your own site. To customize them, just go to Feedjit and customize the current widgets, and then grab the query string for those and append them to the new widgets javascript URL.  If you do install these, I’d love to get your feedback, so email me at mark-at-feedjit.com.

Mark.

…especially when you’ve been quoted on the front page of a major newspaper. (The Seattle PI)

Congrats to Brian Dorsey and his team who built and launched Tagmindr in 6 hours. Brian and I have chatted via email several times this week and we still haven’t met. Hopefully this week at Seattle Tech Startups meetup.

Mark.

ps: Here’s the original blog entry that launched Feedjit. Here’s John Cook’s article online.

There’s an interesting conversation thread going on at publishing2 regarding Facebook apps. This quote from Dave McClure, who I know from the days when my job search engine used to compete with his job search engine [and who I have the greatest respect for].

Interesting use of one data point to provide the proof for a wide-ranging empirical assertion.

i guess as a comparison, i can give you 90 other data points from students in the Facebook Apps class i’m teaching at Stanford this fall who would offer a contrary perspective. they have formed 30 teams of 3 to build apps and learn about using Facebook as an launchpad for startup entrepreneurship. i doubt any of them feel like they’re wasting their time, as you suggest.

I’d love to know the assumptions behind Dave’s course.

10 Million profiles have your app installed (See the graph on my recent blog entry)

Each of them gets 3 hits per day on average.

A CTR of 0.04% to your own website (based on FB’s ad CTR)

= 12,000 pageviews per day or

360,000 pageviews per month.

At $10 CPM (adsense is more like $3) you earn a grand total of:

$3600 per month with 10 Million Facebook users using your app.

Just for fun, increase that to 30 hits per day per profile and you’re still only earning $36,000 per month. Hardly a VC worthy investment.

But what really bugs me is the strategic implications of being a remora to the great white shark that is Facebook.

To me, building a business around a Facebook app feels more like being an eBayer both in terms of the hard ceiling on your businesses scale and the total reliance on the facilitator you’ve hitched yourself to.

What am I missing?

Tim Oreilly has done some data crunching and has a graph showing the distribution of Facebook application use. He describes it as a long tail which is misleading because there is no way for any one business/entity to aggregate the long-tail into something useful - besides Facebook themselves which illustrates how smart their platform play is.

But it’s great data. Combine this with the average ad CTR of 0.04% on Facebook (which is relevant for widgets because they’re subject to the same CTR’s) and it makes one wonder why anyone would want to play in that sandbox.

The other interesting thing about the graph is that it’s a who’s who of Facebook:

One of the first books I ever read and one that has had a profound impact on my life was Jonathan Livingston Seagull. If you recently left the flock or are planning on leaving, I recommend you find a quiet corner and draw strength from Jonathan.

I think that this book should replace Sun Tzu’s ‘The art of war’ as required reading for entrepreneurs. I was recently watching a video of Sergei Brin giving a talk at Berkeley. One of the students in the audience asked him if thinking about competitors like Microsoft keeps him up at night. He replied that he doesn’t worry too much about competitors and that thinking about the incredible opportunity that he has is what keeps him up at night.

Here is the first few paragraphs of Richard Bach’s Jonathan Livingston Seagull:

It was morning, and the new sun sparkled gold across the ripples of a
gentle sea. A mile from shore a fishing boat chummed the water. and the
word for Breakfast Flock flashed through the air, till a crowd of a
thousand seagulls came to dodge and fight for bits of food. It was another
busy day beginning.
But way off alone, out by himself beyond boat and shore, Jonathan
Livingston Seagull was practicing. A hundred feet in the sky he lowered
his webbed feet, lifted his beak, and strained to hold a painful hard
twisting curve through his wings. The curve meant that he would fly
slowly, and now he slowed until the wind was a whisper in his face, until
the ocean stood still beneath him. He narrowed his eyes in fierce
concentration, held his breath, forced one… single… more… inch…
of… curve… Then his feathers ruffled, he stalled and fell.
Seagulls, as you know, never falter, never stall. To stall in the air
is for them disgrace and it is dishonor.
But Jonathan Livingston Seagull, unashamed, stretching his wings
again in that trembling hard curve - slowing, slowing, and stalling once
more - was no ordinary bird.
Most gulls don’t bother to learn more than the simplest facts of
flight - how to get from shore to food and back again. For most gulls, it
is not flying that matters, but eating. For this gull, though, it was not
eating that mattered, but flight. More than anything else. Jonathan
Livingston Seagull loved to fly.
This kind of thinking, he found, is not the way to make one’s self
popular with other birds. Even his parents were dismayed as Jonathan spent
whole days alone, making hundreds of low-level glides, experimenting.

The book is published in its entirety here.

One of the better ad images I’ve seen lately - courtesy of seaspace.cn:

The biggest freshwater fish caught by a woman record in the UK has been smashed. This old man was safely released after having his weight verified by the lake bailiff.

Details on “A Welsh View”.

Feedjit made Techcrunch this evening and Michael has installed both our widgets on the blog entry, complete with his own personal color scheme “Arring Town”. We also rolled out a new version that adds a pretty cool (if I say so myself) ajax widget customization interface. Now you can choose your own colors - a much requested feature - and you can adjust the traffic feed width.

If you have any questions, bug reports or comments, please don’t hesitate to email us at support@feedjit.com or just leave a comment here.

Here’s a clip my bro pointed me at from Mike Patton’s new project Peeping Tom. If you have a very wide range in musical tastes (from BeeGees to Tool) then you’re going to love this:

The thing about running a widget business is that you serve as many web server requests as all your users websites, combined. And if one of your users get’s Dugg or Slashdotted, you get Slashdotted too.

After I launched FEEDJIT on Thursday (5 days ago) the traffic started picking up Friday and by Saturday morning my server was groaning under the strain. Some of the highest traffic blogs were Japanese (there are more Japanese bloggers than English) and by mid-morning the Japanese were going to sleep, so that gave me a welcome reprieve.

The first thing I did was reduce Apache’s KeepAlive timeout to 2 seconds. KeepAlive’s let clients hang on to a connection which someone else could be using. If a client uses keepalive properly then it can give you a nice performance boost, but set the timeout low so slow clients don’t waste server resources.

Then I added HTML caching for the widget serving routine using Perl’s Cache::FileCache. This gave me a huge speed increase but the stats on our widgets were 1 minute delayed - and that sucked.

By Saturday night I’d rolled out the new caching code and the server was a lot faster, but I knew it wouldn’t work long term and non-realtime stats for FEEDJIT was not an option.

By Sunday I was getting 40 hits per second and rising and the server was groaning again. I had to make some fundamental changes to the way the app was architected. The old mod_perl2, MySQL and Apache2 combination wasn’t going to cut it.

So I basically redesigned the data storage routines from the ground up. I moved from mysql to a home grown data access method.
I can’t tell you how gratifying it was when I rolled out the new code last night and watched the server load average drop from 2.5 to 0.3 (unix load where 1.0 = 100%) and hold there as our traffic continued to rise.

We have several high traffic blogs now and our busiest blogs generate around 1.5 widget loads (pageviews) per second. I’m confident that if for some reason TechCrunch added us tomorrow, we’d easily handle the traffic without breaking a sweat.