From Istanbul To Sand Hill Road

The Most Significant Thing Google Did In 2005

I finally found some time to read the 2005 Google annual report.  I recommend any investor to get in the habbit of reading such documents (10-K's, 10-Q's), because they give you the real story behind a business, not to mention a birds eye view of what's going on that's not easy to see following day to day news.  If somebody had asked me, "what's the most significant thing Google did in 2005?" I would have said a whole bunch of things, all of which would have been wrong, but reading the annual report makes clear what the right thing is.  I'll get to what I think is the most important thing later, but first I want to mention a few things that caught my attention.

1) In 2005, Google bought 15 companies for a total of $85M.  That's less than $6M per company, and a scary number for VC's who are investing in Internet companies with hopes to sell them to Google.  They'd better be really, really capital efficient :-)

2) Top competition is Microsoft and Yahoo.  No surprise here.  But we know what Microsoft is doing about is, you may have seen Steve Ballmer's "advertisers, advertisers, advertisers" video.  They want to get a hold of all the advertisers and choke Google there.  Again no surprise.  What's surprising to me is that nobody know what Yahoo is doing?  Are they doing anything?

3) Google did $6.1B top line, $1.5B bottom line, but increased it's cash position from $2 to $8B.  They did that by raising $4B in the secondary offering.

4) They spent $483M on R&D but spent $1,417M on property and equipment, not counting depreciation.

Point #4, makes me conclude that point #3 is the most important thing Google did in 2005.  It's raising that $4B.  Why?  Because it gives them a war chest that Yahoo does not have.  In 2005, Yahoo generated $600M in cash, and Google generated $6B.  This could spell disaster for Yahoo, if Google starts to force them to spend capex at a rate that they can't keep up.  While Yahoo tries to keep up with Google coming up with products, they may lose the game if they can't bid for the data centers like Google can.  They may lose the war by losing the hardware battle, not the software. 

I have a lot of respect and admiration for the products Google is building, but if somebody asks me what's the most important thing Google did in 2005, the answer is the secondary.  Obvious as it sounds, I bet it wouldn't be the first thing most people say.

What Is Society Thinking?

Google Trends is a new beta product from Google that graphs the relative frequency of a search term over time.  Since it compares the number of searches for a particular keyword  to all the searches (in a representative subset) it is a proxy of what the population overall is searching for.  Once out of beta, this could be pretty interesting, after all it tells you what our "society" is thinking of.  I played with it for a while today, to see what society is thinking of, really.

Some things are obvious, but still pleasant to see.  There are a lof searches for 'Halloween' every October and it drops like a rock shortly after.  The comparison of 'shorts and sweater' show the cyclical nature of the keywords; society is either searching of one or the other, but not both.  It's either winter or summer.  We are looking for shorts in June and sweaters in December, so I guess the northern hemishpere is winning :-)

This is all nice and fun, but I wanted to find trends that actually tell us something about where society is going.  What big trends could I glean?  Of course there are a lot of things that are popular today that show a nice trend upwards, say 'Web 2.0' for example.  That doesn't interest me much because it could be a hype.  What I kept searching for is a steady decline in search frequency.  What are we caring less and less about?  This turned out a lot harder to find.  Most declines I found are blips in frequency, like 'iraq' , or cyclical like 'summer camp'.  The two I found are 'television' and 'newspapers'.  Then I asked, what are TV and newspapers replaced by, and so compared 'television vs. Myspace' and 'newspapers vs. blog'.  Both results are very entertaining to see, check them out.

I am very curious if others can find more interesting trends and counter-trends.  Please let me know.

On an aside, I did use Google to compare my favorite soccer team Galatasaray, to their lesser known, niche rivals in Turkey, and long behold I saw a very satisfying trend indeed.  The obvious is confirmed by Google here.

Cometa Done Right?

Cometa was a failed venture backed startup that aimed to build a nationwide network of WiFi hotspots and wholesale the network capacity to ISPs.  They were a WiFi carriers carrier.  It was a noble attempt to make broadband available to the nation, but eventually failed.  Main reason of their failure was coverage issues that resulted in low end-user uptake.  This was at a time before 802.11g, 802.11n, and most importantly before Wimax. 

Today on VentureWire, there was news of a Venture backed startup (Redpoint, CRV, Kleiner) named M2Z which applied to license spectrum from the FCC to provide nationwide broadband, answering the call of President Bush to make it happen.  The company is founded by 1st class technologist Milo Medin (@home) and John Muleta (FCC).  Their proposal is to get the nationwide spectrum free, with the condition of making service availbe to 95% of the US, free to consumers for under 500Kbps (ad supported), and offering the government 5% of revenues for anything they sell better than 500Kbps.  My hats off for this bold vision. 

They have a better shot at succeeding because, the cost of building a nationwide network isn't as expensive as it was long ago.  I've written about this many months ago here.  Wimax cost per home passed numbers are under $10.  No doubt, M2Z says they've secured $400M of debt to fund the buildout.  So there is a wireless backhaul option that wasn't cost effectively available to Cometa.  On top of that, coverage with 802.11g, or n is a lot better as well.  So this is the right time to build this kind of network.  Since it is wimax, it does require a CPE, the report says will be reasonably priced.  Milo Medin started @Home, and at the time a reasonable priced cable modem was $250-300 (we built it at 3Com).  The pricepoint is understandable, because it will have to have wimax chips, antennas, and most likely wifi on the other end, after all you need to be able to use your laptop with it.

Does M2Z want to sell the service themselves, or wholesale it to others who want a two-way pipe into the home?  If it is the former, you need to add a subscriber acquisition cost on top of the CPE price to see what a hole M2Z has to crawl out of, if it is the latter, their monthly price to the carrier will be a more modest number say $5-10/mo.  My bet is the latter, the Cometa model.

It could work.  The key thing will be the price that they will be able to charge the end user.  Will a local MUNI wifi provider, or Google in SF, make the service free so that nobody would ever use M2Z?  Unknown.  If it is a billion dollars or two, can MSFT or Google build this themselves, and make it free and totally ad supported even at the higher bandwidths?  Also unknown.

Let's do a back of the envelope calculation on the business model.  If M2Z gets $5/mo revenue per paying sub selling wholesale, and operates at 50% EBITDA, that's $2.5/mo to cover the $300/sub investment (assume $275 CPE + $25/home passed+other).  The buildout cost is paid by debt, at 8% interest, the $250 per sub investment costs them $24/year, barely covered by the $30/year contribution margin. 

If they get $10/mo then they can pay off the debt service on that sub and generate about $35 or margin per user per year plus advertising.  This looks a lot better.  If they can get 10M users, this is pretty interesting.  Of course, that's the big if.  What will the alternatives be at the time?  My guess is that there will be a lot of nearly free options already, but who really knows?  It will be close, that's for sure. Since I am not an investor, it's not much of a worry for me on the business side.  On the other hand, it's a bold effort to change the world (as was @home) and I commend the courage and vision of the investors to do it. 

Wireless Ventures 2006

I was fortunate enough to be a panelist at the VentureWire Wireless Ventures 2006 conference today.  The main focus of the event was consumer applications and services.  Since most of my focus lately is on Web 2.0 services and apps, it was a nice chance to see what's going on in the wireless world and compare it to what's going on in the Internet.

After sitting in 10 presentations and talking to entrepreneurs, two things became painfully clear:

1) The ultimate winners are the wireless carriers.

2) There will be a lot of VC money lost, especially money invested since 2005.

The wireless carriers are the ultimate winners because they still control the pipe, by virtue of owning your phone number, the billing relationship and the software that shows up on your phone.  The wireline carries have long become a dumb pipe and have no such power, the IETF took care of that.  In the wireless world, if you have an app and you want it to be featured on the carrier deck, you have to pay them 30-40% off the top.  If you want to go off deck and sell a ringtone, you can do that, but premium SMS is the preferred and maybe only feasible way to pay for it, so again the carrier gets its 30-40% cut right off the top.  Even if there were no transaction fees, and the app was free and ad supported, if you want to be on the deck, you have to share advertising revenues.  So no matter what you do, the wireless carriers, get their pound of flesh.

Let's take a classic casual arcade game, Boulder Dash (which is a classic, and one of my favorites, I played it when I was a kid, now my daughters play it), and compare the experience of buying it for your PC to buying it for your handheld.  Type the word in Google and it will tell you where to buy it.  Their web site takes credit cards and probably paypal, and you are done.  To get the wireless version, there is no good way to find it, and if you are lucky enough to find it then you have to pay through the carrier and they get their cut.  Carriers have a lot of power in the wireless world and they know it.

Why do I think VC's will lose money?  I heard there are some 10,000 wireless apps all trying to get the top 5s slot in a carrier deck.  Not to mention they have to give a cut to aggregators.  These infrastructure players are gettting bigger and stronger too.  They are carrying hundreds of titles and have now, long-lasting relationships with carriers.  Verisign recently bought M-Qube for example.  So even if you invested in a application publisher, or infrastructure company, you have serious competition and taxes to pay.

If you were a VC investing in mobile apps 2-4 years ago, you did the right thing.  There was a lot more open space.  Jamdat, Mobile 365, M-qube, all got funded at that time, two of them exited.  If you were investing in 2005 onwards, I think you have a tough time ahead of you.  Whether you are an app or infrastructure company, you are now facing powerful carriers, big publishers and big infrastructure players.  Small publishers, however strong a niche they think they've carved for themselves, will have a hard time getting their titles to sell.  It's simple, you are one of many trying to sell to a few customers.

So what's left?  Where are the opportunities?  The one place where I see room for innovation is in ad supported, free, off-deck applications.  In this case, there is no carrier tax paid.  Aggregators could be avoided too.  All this is good but the big question is, how you acquire customers.  For the ad supported model to work you need to have a lot of subscribers, viewing lots of ads (and don't count on $30 CPM ads either, that's temporary due to lack of supply, assume $1-3 CPM ads realistically).  You can't get subs using search marketing unfortunately.  So it's up to the smart entrepreneur to innovate in business development and marketing.  What kind of partners can bring me the most users?  What kind of features will make my app distribute virally?  Companies that can crack these questions are interesting.  Best of all, there were a few such companies at this show, but I can't tell which ones.