Recently in Numbers Category

I usually don't get to write about the events I attend, but with two events in a row (Facebook's F8 and Venture Beat's MobileBeat 2008) and two more to come (TechCrunch's Mobile Web Wars Round Table and TechCrunch's Annual August Capital Meet Up) I need to organize my notes -- and why not do it here on the web.

F8 was a waste of time in general - Mark Zuckerberg's keynote was fine but the next talk that followed him on stage was long and mostly just asking-for-applaud.  it became boring rather quickly, and the different tracks did not offer much.  I switched between sessions in hope of finding useful talks, but at least to me the time could have been better used.  I left early.

MobileBeat was a different story.  There were four panel discussions, and pretty much all of them offered useful information.  The panelists knew what they were talking about.  I for one learned quite a bit.  I'm sure different people left with different takes -- here's mine.  These are not necessarily related - just notes from the event:

• A lot of talk about iPhone.  However, people did mention a few times that internationally the number of iPhones is insignificant compared to other handsets.
• Nokia has a 70% penetration in India.
• The problem with Apple's iTunes App Store is that it's another walled garden.  And the same old pattern is happening there:  A small number of apps show at the top of the list, they get more downloads because they are at the top, and they keep staying at the top.  Vicious circle.  Few apps will find a chance to be "discovered".
• About 6500 apps exist for BlackBerry today.  About 70% of these apps are vertical.
• Unless you have a specific network and customer line up, being vertical makes apps less attractive to the VCs since there's already a small percentage of people who download apps, and for a vertical app this makes the number of downloads (hence growth) even smaller.
• For the same reason VCs prefer to see apps that would work across all platforms, not just iPhone.
• At the same time, a number of companies such as Loopt has found iPhone a good platform for the initial launch - call that your beta.  Do it on iPhone because development is short and download problems are not huge.  Then study the reaction of your users, who are usually the early adopters.  If that works out well, then think about releasing the same app to other platforms such as Windows Mobile or Symbian.
• Despite all the talk about how Apple is eating AT&T's lunch with the app store, AT&T is not much interested in the app revenue.  The revenue from the customer, including the voice, SMS and data plans, dwarfs the app revenue.  So they prefer to focus on the infrastructure and the "network", not the apps.
• VCs are less interested in replicated what's available on PCs on the mobile devices, that's usually done by larger companies anyway and is hard to compete with.  They look more for things that would have not been possible if the devices were not mobile.
• Games are the most popular apps on mobile devices.
• Among interesting topics to mobile VCs are edge issues (such as acceleration), identity management, and discovery of content/apps.
• iFund seemed to be pleased with the investments so far, and they seem to be ready to expand their fund if they see more iPhone opportunity.  Before iFund they have practically never seen a "successful" mobile investment.
• Numbers are amazing (or I should say scary):  In all these years only about 143 mobile M&A transactions have happened in the market, of which less than 20 are more than $200M, and less than 35 are more than $100M.  This means little success in mobile VC investment.
  
• Advertisement is still considered the biggest way to monetize.  However, brands do not prefer deep-inside-the-app advertising, because the user experience does not allow the user to easily "click through" to the target site - user usually does not (or cannot) leave the app.
• Referral fees are also a good business model.  It's possible, for example, to have an app on iPhone that refers the user to a song within iTunes.  If the user buys the song the app gets a cut.
• Also using Virtual Goods as a monetization strategy came up a few times.  Interesting.
• MySpace has about 1.7M users on their mobile web portal.  Facebook seems to be about the same.  These are US numbers.  MySpace also has downloadable apps, but they did not disclose numbers.  I guess if download was in the same ballpark as mobile web users (or more) they would disclose, so it should be considerably less.
• Some interesting numbers in terms of mobile usage in the US:
• Number of subscribers:  232M in '07 ---- 255M in '08
• Those using SMS: 140M in '07 ---- 198M in '08
• Those using MMS: 77M in '07 ---- 123M in '08
• Those with data plans: 74M in '07 ---- 95M in '08
• Of 95M users with data plans ony 40M actually use Internet services, of which 30% are on smartphones.  4% on iPhone (part of the 30% smartphone), but the iPhone users' usage is 5X the average smartphone user.
• The discussion on whether one should write the app for mobile web or as a downloadable app did not coverage to a specific black and white answer.  The consensus was that you should look at your audience.  For example, if it's a business app you would probably want to write an app for BlackBerry.
• AdMob's CEO was emphasizing that advertising on mobile is already happening.  They are serving 3.5 Billion ad impressions (sorry - can't remember if it was "per day" or "per month" - but the number is too huge anyway) internationally.  The CPM is $10-$30 in the US.  A lot lower internationally.
  
• Forget about Java development for handsets in the US.  The download issues are just too much hassle.

I'm sure there's more, but this is most of what I remember from my notes...

"At a hardware BOM and manufacturing cost of $173, the new iPhone is significantly less expensive to produce than the first-generation product, despite major improvements in the product's functionality and unique usability, due to the addition of 3G communications... The original 8Gbyte iPhone carried a cost of $226 after component price reductions, giving the new product a 23 percent hardware cost reduction due to component price declines."

Read the full article at iSuppli's Preliminary Study.  The image shows the BOM from the iSuppli article. Click on the image for larger version.

Running on less than 300 servers, Wikipedia boasts some huge numbers:

• 50,000 http requests per second
• 80,000 SQL queries per second
• 7 million registered users
• 18 million page objects in the English version
• 250 million page links # 220 million revisions
• 1.5 terabytes of compressed data

Read the full article at A Look Inside Wikipedia's Infrastructure - Data Center Knowledge.

Some pretty interesting statistics about the mindset of teenagers when it comes to shopping and online life:

• They prefer to shop in a store (82%) to shop online (18%)
• They prefer to IM a friend (54%) over calling (46%)
• They would rather get their locker vandalized (63%) than their homepage (37%)

No comment!

I have followed Jingle's Free411 service (1-800-FREE-411) since they are a voice service like Frucall's initial voice based comparison shopping.  They have done an amzing job, specially given the competition from Google (1-800-GOOG-411).  At this point they have about 6-7% of the US directory assistance market, and they are processing about 20M calls per month.

The company had raised two series of financing before and the valuation was around $150M in 2006.  They just announced a series C round of about $13M with existing investors.

The amount is low compared to the valuation and their series B, so I'm not sure what exactly is going on.  The company has just reached profitability, and their market share and call volume is great.  I would expect a larger amount for aggressive growth and capturing more market share.  I hope this is not a down round for them for unknown reasons.  Most probably their estimates about their expenditure have been off, and they are running out of money sooner than they think so they are doing an internal round.

TechCrunch has a nice list of the VC deals in the US Web 2.0 companies:



The table shows the last round of financing.  Some more data about total money raised can be searched on Private Equity Data Center

Some good data about web usage and penetration at TechCrunch.

These numbers will give you an idea about how much processing Yahoo! and Google are doing to index to web in order to let us search it...

Here comes the latest numbers on visitors of personalized home pages, reported by ComScore via TechCrunch.  Yahoo! shrinking a little bit, Google growing big time, and bottom line looking pretty bad for the leading start ups in the field, namely NetVibe.