Amr Awadallah Blog

About two years ago I wrote about switching from Netflix to Blockbuster, well, now I am switching back to Netflix.

The reasons for the reversal:

1. Netflix now streams movies to my PC (in home office), Xbox 360 (in living room), and PS3 (in bedroom) :). I only ask that they expand the online streaming selection to include newer movies, even if I have to pay an extra per movie fee for that (I don’t like to pay XBox Live or Playstation Network $5 to see a movie, that is such a rip off)
2. Netflix now is much better at shipping out blockbuster titles without a long wait (I actually timed Blockbuster vs Netflix head to head on a number of popular new movies, Netflix always shipped out first)
3. Netflix’s web user interface is more intuitive and much snappier than Blockbuster’s, especially when it comes to editing the movie queue and moving things around.
4. Last, but not least, previously with Blockbuster when I dropped a movie at the physical store, I could pickup another movie from the store then they immediately shipped another movie from my queue. They changed that system, they now don’t ship another movies from the queue until I return the movie I picked from the store. Furthermore, they now limit the number of in-store swaps that you can do. This was actually one of the main reasons why I started looking at Netflix again.

Bye bye Blockbuster.

— amr

This is pretty much your last chance to signup for Hadoop World, it is happening this Friday (Oct 2nd) in Newyork City. The registration prices are higher now since these are the last few days, but if you are interested in going then shoot me an email, I have a special deep discount code for “friends of amr”.

Below I highlight a number of news/blog articles that covered some of the use cases that will be talked about at Hadoop World:

• “Hadoop buzz continues to excite the cloud”. CNET, Dave Rosenberg, September 4th, 2009.
• “What’s the New York Times doing with Hadoop?”. InfoWorld, Savio Rodrigues, September 11th, 2009.
• “Telecom operators get the scoop on Hadoop”. Telephony Online, Rich Karpinski, September 15th, 2009.
• “Q&A: Visa dips a toe into the Hadoop pool”. CNET, Matt Asay, September 17th, 2009.
• “The Tech Behind eHarmony”. PC World, Eric Lai, September 19th, 2009.
• “Hadoop World Shows Diverse Uses”. SD Times, Alex Handy, September 21st, 2009.
• “Cloudera and Facebook Shed More Light on Hadoop Integration”. TechCrunch, Leena Rao, September 23rd, 2009.
• “Welcome to Hadoop World”. Network World, Jon Olstik, September 24th, 2009.
• “FeatherCast – Aaron Kimball on Hadoop”. Apache Feathercast, Rich Bowen, September 24th, 2009.

I look forward to meeting you there,

— amr

First, I can’t believe it has been three months since I last blogged, all I can say is that a startup will suck the life out of you 🙂

I still see many startups creating free online services and expecting them to generate hundreds of millions of dollars from advertising. I might be stating the obvious here, but that will only work out if these services get tens of millions of users, with those users returning to the service almost every day, and that will only happen if the service meet the essentiality criterion. Essentiality is the litmus test that differentiates a life-style business (e.g. a super popular blog that makes a few million $/year max) versus a venture business (i.e. a business which generates hundreds, if not thousands, of million $/year). By essentiality I mean both the ability to amass a large population of users, but become necessary to these users such that they need to visit the service almost every day (if not many times a day).

For example, here are some essential online services:

• E-Mail/Chat/IM: It is essential that you communicate, almost continuously
• Search: We all need to find stuff many times a day
• News: Almost everybody needs to know what is happening. Facebook is news about your friends, it is communication too.
• Music/Movies/TV/Videos/Games/Sports: A decent population needs their daily entertainment “fix”, however very few destinations are able to aggregate a large enough audience due to content fragmentation

I was then going to list the many services/sites that I don’t think will make it due to failing the essentiality criterion. Many of these services will collapse due to fact that they are small/life-style businesses but their VCs expect them to hit venture scale (the ones structured as small businesses will live and prosper). But, to avoid alienating such businesses (since some of them are potential customers for Cloudera) I decided not to list them here :). That said, the common pattern among them are: either they have a lot of users but are not essential to any of them, or they are useful but to a very small group of users, in both cases this means they will not be able to amass the huge volume of pageviews required to make venture-scale money based on advertising.

Cheers,

— amr