Friday, April 11, 2008

web 2.0

funny I came across this blog.. it was written in 2006 and talked about web 2.0 fatigue.

"But reading TechCrunch, Mashable and so on has lately become less exciting for me. The Web 2.0 list is massive - and there are many people bringing out similar sites and services."

funny it seems nothing has changed since then.. but for I guess youtube and twitter.. :)

Wednesday, October 31, 2007

The buzz just got intense. Google Open Social, Facebook SocialAds

Three things happened this week. and the impact of those three rumors/ announcements might be so huge that it has the potential to change technology landscape will change for the next few years. Of course, it is possible that might not happen (remember Google Base) or it might take years before it makes an impact (remember wikipedia).

So they are..

Firstly Google will launch OpenSocial platform tomorrow. Nobody else can cover the technology impact of the announcement better than Marc Andreessen.

Secondly, Facebook plans to launch Social Ads. I believe this is a powerful idea, an idea which I myself have toyed with, so I guess I am biased. But I do believe this is one of those ideas which will have impact on advertisement payoffs, better targeting and even change the perception users have towards ads. If the ads become personal, more relevant and useful, I will not be so averse to them.

Thirdly, Google planning to work with Verizon (and/or Sprint) to be their search engine. I bet by next year, they would be their mobile OS as well.

Now thats a lot of news on several fronts. That also means that we are finally in those bubbly and feel good years for technology companies after years of dullness. and its not bad. Because technology and developers thrive on emotion. Its easy to get motivated when the market is strong and everyone can dream of riches ahead. Thats when the most risk is taken and thats when ideas even from small guys get experimented with.

But coming back the theme, what is this all about. Is it about "Do no evil"? Or is it about dominating technology headlines. Besides both of the above, it is mostly about being able to "own the data" that will lead to "better targeted advertising" in consumer world and "better customer leads" in enterprise world. That is the real treasure that Facebook/Microsoft and Google are fighting over.

Friday, October 26, 2007

SPAN SUMMIT LIVE: Scaling Facebook Apps

Pretty good pointers for developers of facebook apps around Scaling Facebook Apps

By Mark Mayo; CEO, Joyent

How to Scale

  • Vertically (db)
  • Horizontal (page generation) – stateless page + segregated inserts
  • Agile Infrastructure + Dev – Optimization takes priority over features as soon as app goes viral. Users arrive fast and furious


Example of “Are you normal” by Kinzin

- They had their own infrastructure from independent website

- 1sst facebook app

- but used joyent

- Traffic: 150k yesterday, 200k today

What they did in the app

  1. think about reengagement
  2. First 50k users are most difficult. Advertised to get those users
  3. got to recently popular page. Got peaks when they appear on recently popular pag

Their Story

Started Small

  • started with $15 shared hosting
  • chose $45 accelerator
  • 2x125 (1GB) for launch
  • preconfigured and tested load balancing

First Recently popular page

  • 10K users
  • db was 100% cpu
  • not enough mongrels to handle db connections
  • what they did
    • split and clone stateless app server container
    • double database memory
    • tripled performance in 2 h

Its always database

- look for slow queries (obvious but overlooked, untested)

- find offending code, hand optimize sql (do it)

- memcache everything (no other option)

- run a memcached per app server

- most apps don’t need to go to multiple db (very expensive option)

Second recently popular

- 50k users

- much better caching

- just one slow query

- added 2 more app servers (4x now)

- db lagging on I/O now

Recently popular #3

- already dedicated disk spindles for data and logs

- 100k users

- no problems, db cpu under 15%

What happened: went totally viral in 2 weeks. Very little time if your app goes viral. Have a strategy.

Next Steps

- mysql replica

- split read/write queries in code

- add appservers as needed

- metrics indicate another 3-4 app servers to reach 1M with capacity to spare

Lessons learnt

- you only get hours on recently popular page

- must have plan to add capacity on fly in tens of minutes

- plan on how to attack optimization in ur code

- test with more than sample data to find those slow queries.. its always db


- to load test.. write script to add users

SPAN SUMMIT LIVE: Facebook platform by Ami Vora

Facebook platform by Ami Vora (Facebook)

Talking about traffic stats (hovers between google, ebay, 5th most visited site)

Social graph

Photo tagging, awesome ability to reach out to only those who matter

same old stuff..

Now comes good stuff..

Platform offers

  1. mass distrivution – content, invitations, applications, anything oif value
  2. deep integration – if Facebook can build it, so can you
  3. new opportunity –
    1. Growth – access to millions, viral diftrivution
    2. Engagement – social context, 50% retuen users on website
    3. Monetization- freedom

i. Ad n/w

ii. Cross promotion

iii. Institutional funding

  1. Facebook fund
    1. low barrier to entry
    2. small no equity grants
  2. Advice from facebook
    1. engaging content, focus on social. Focus on user.
    2. use integation points well
    3. develop and iterate.. why will user come back. Low investment, quick feedback, iterate.
    4. intelligent promotion
    5. incorporate privacy
  3. This is just the beginning. Think long term. Same users, so facebook believes success depends on each others success.
  4. Facebook found that good performance have dramatic impact
  5. Q&A
    1. Micropayments , Monetization – two approaches. Either Facebook develops or 3rdparty develops.
    2. Facebook will leave a lot of money on the table for 3rdparty
    3. Best user experience is not possible if Facebook tries to own everything. It is upto users.
    4. They are thinking about enabling news feed stories about applications even if user has not installed but friends have.
    5. FACEBOOK will change utlity (feeds etc) based on what works for users, what works for what kind of apps.
    6. The game is not over. Slide and Rockyou did good with opportunity. Platform matures, applications will change.

SPAN SUMMIT LIVE: Virality Session for facebook apps

Panelists: Jia Shen (Rock you), Dave McLure, Keith (Growing Gift), Mark Kantor
Moderator: Susan Wu (Charles River)

Points made so far:

  1. Virality is: Conversion rate which gets 1 user to invite x. where ideally x>10, but definitely should be > 1
  2. Popular idea today: Send a message.. don’t tell them what it is. Or have a gift, get them to click before they can see it.. or make them wait a few days.
  3. Viralness is different from stickiness.
  4. Active users include app additions. Track how many are returning users vs new users
  5. You can wait to reach out to users who have added the app after rolling out the app
  6. Focus on your active suers. Active -> initial installs and invites following that. Passive -> ocassional users
  7. Ease of development allows you to take a chance.. many developers
  8. Missing features: Discovery of apps. Best way to do that is to see my friend using the app
  9. Get yourself on the minified
  10. Get them to send targeted invites, not just random invites
  11. Analytics: You get to track from beginning to know what to fix.. notifications, clicks, invites, engagement.
  12. First 10, 20 clicks are most important in understanding if the app will grow
  13. Then try out A/B testing. Do Mathematical modeling. Multi variant testing. Grab demographic data and then see which ones do what. Per install, per click what happens
  14. Dave rants again about Acquisition, Retention, Monetization
  15. Humor: Susan said.. is developing facebook apps is like investing in Chinese stock market J . Dave: Big market, opportunities
  16. Graphics in feed. Call for action in the profile
  17. Copy the techniques from slide/rockyou
  18. Use fbnames in the actual message..
  19. Any interactions with multi user .. fan out effect.. feed prioritization
  20. Invite optimization: in new system. It is harder to track, can send 20 (average invite per user= 14). User has to pick. Dave says that’s hurting developers.. not sure why facebook is limiting. Keith: fb raised the bar that user has to be passionate abt app now.. because they got to select each user. Dave: danger.. fb users might be getting tired of spamminess, hence the move. Fb profiles are beginning to look like mysoace profiles.. –ve effect
  21. Jia: invite rates are tailoring down.. User profile is their social capital. In the beginning, it was converting crazy. Last 3 months, things have been same
  22. Mark (kickflip): Strive towards 100% acceptance, hide the message, no single sign on experience, so they have to add the app
  23. Major stats attention: what do they care about

a. Mark: small apps, so focus on growth. Daily/weekly Time on canvas page is good metric for ad monetization

b. Keith: active users/day. engaging, what animals popping out of eggs..

c. Dave: over emphasize on metrics.. cpa might be better than impressions on facebook

d. Jia: Numbers are bragging right. Actual stats which you care about are different.

e. Susan: engagement metrics should be different for different apps.

f. Jia: Uninstall rates: 10-20% uninstall rates.

  1. Top things to take care about: Notifications, profiles are better for growth for new apps.. feed still up in the air. Optimizing notification to friends. Make invite personal. Notification should call for action. Channels: invite, mini feed, name of app, profile action (default profile action.. what other users see when browsing). . use sexiness
  2. Kiss me, dodge me: new apps by Stanford Dave’s class
  3. Spam tune. Never send more than 1 notification per user per day. Wait to unblock. Notifications sit for ever..
  4. Audience: primarily N. America, UK. Facebook just announced they are localizing
  5. Analytics tools used: Google analytics, page tagging. Otherwise custom sql statements, log watch. Dave: Google anaytics.. introduced log events for rich media
  6. Other SN: Reprioritize based on developments
  7. Team structure
    1. Mark: 5 full, 2 partime, 1 business, 1 ad
    2. Jia: 20 ppl, 5 business
    3. Dave: 25 teams of 3
    4. Keith: 2 full, 3 partime, 2 developers.
  8. Focus on creative side of apps. If you are not equally excited, that’s too bad.

Sunday, October 14, 2007

exception - for the man in the arena

I don't usually do this.. that is refer blogs I read somewhere unless I have some thoughts of my own to add to them.

but for this post (surprisingly its from Mike at TechCrunch. which is fine I guess, its just that I never thought I would be making an exception for Techcrunch post), I will make an exception. enjoy!!

generation gap on facebook

I made an observation in my last post - "The 25+ browse more but participate less while 16-25 engage facebook using it as their predominant social communication and entertainment tool." purely based on what I saw around me and on facebook.

Today it was validated by the mighty NY Times. "To some extent, a generation gap is already apparent in the Facebook population, said Mr. Shirky, of N.Y.U. Younger people will use it more naturally and differently than older folks, who for the most part will see a Facebook page as something like the dreaded Christmas letter, with its prosaic updates on one’s life events, and less the sense of “living your social life online, hammer and tongs,” the way younger people tend to.

That is only natural, he said, because “People our age are going to find uses for the tool that have to do with the maintenance of life already in process, rather than making one up out of whole cloth” — in part, he added, because “our social lives are more boring.” And that, he said, is only logical: “We’ve made the mistakes we’re going to make, God willing, and we’ve settled down.”

What driving this.. peer pressure, word of mouth, gossip, curiosity etc. Amazing that the percentage of active members who are over 25 years old and out of school has risen to some 40 percent of the overall population of about 45 million on facebook. And many of them signed up to check it out and willing to hang out.. nice!!