It's an oft-stated truism that 90% of all new business ventures here in the USA fail within the first two years. According to the Small Business Administration, what kills most small concerns is under-capitalization -- which is to say that they run out of money, either before they can turn a reliable profit or by over-extending themselves in an effort to expand.
Around these parts, it's the first one that gets 'em, pretty much every time -- and the failure rate for new businesses here in Mariposa is nothing short of appalling. They perish like mayflies and, more often than not, their final exit occurs in the dead of winter.
The problem is that Mariposa's economy is almost exclusively dependent on tourism, but the vast majority of our tourists visit only during late Spring, Summer and early Fall. The rest of the year, local enterprises must depend for their livelihood, not on the kindness of strangers, but on the custom of year-round residents -- and there are barely 17,000 of us in the entire county.
That's an awfully small customer base, and it's precisely why so many start-ups starve to death here, waiting for the tourists to return. If they're too dependent on income from our visitors, they just don't survive past the Winter Solstice.
And that same kind of inexorable logic is what I think will kill 3G.
Castles Made of Sand
There's a reason why economics is known as "the dismal science". The laws of supply and demand, income and expense, profit and loss are downright Darwinian. And a balance sheet doesn't care how many hopes and dreams you've invested in your business, because those things have no effect on its bottom line.
To my mind, that's exactly what stands in the way of 3G's success -- its business model is based more on wishful thinking than on market realities. The way I see it, in terms of spectrum license fees, infrastructure buildout costs and service pricing models, the wannabe 3G moguls are simply kidding themselves into rapidly-impending bankruptcy.
Let's start with the ridiculous sums that licensing spectrum will command. In Britain, all of the major incumbent cellular carriers were successful in their bids for 3G spectrum -- but collectively BT Cellnet, Orange, One 2 One, Hutchinson 3G and Vodafone paid in excess of $35 billion for those victories. German carriers, not to be outdone, then threw a whopping $46 billion or so at their government's auctioneers.
It will take the British carriers close to 15 years to recoup their investment in spectrum alone -- assuming that they can manage to retain the current penny-per-kilobaud pricing structure they enjoy on existing service -- and those $35 billion British licenses expire in just 20 years. And all five companies were staggering under massive debt loads even before they got caught up in the auction fever.
The situation is every bit as bad in Germany, where six operators will somehow have to find enough credit -- because they, like their British counterparts, have essentially zero cash -- to build out the infrastructure they'll need in order to exploit the licenses for which they've agreed to shell out over $46 billion.
And neither government appears inclined to offer rebates to the suckers..er, that is "winners"..in their respective lotteries.
No Sugar Tonight
Leaving aside, for the moment, the estimated $100 billion or greater worldwide price tag on the new construction -- antennas, transmitters, switches and so on -- needed to make this brave new cellular world run, there's an even bigger problem looming for the 3G folks: their projected market. To put it bluntly, the pot of gold they're chasing probably doesn't exist.
To begin with there's room for considerable doubt that users are going to be willing to fork over five or more times what they're currently paying for the privilege of listening to CD-quality music on their cell phones or watching video clips on two-inch screens. After all, the explosion in MP3 players already gives listeners the first capability without forcing them to incur per-minute charges and portable TVs -- content for which is also free -- which currently address the second aren't exactly what I'd call ubiquitous technology.
And early adopters, enthusiastic evangelists though they may be, simply don't a mass market make.
I can't help but think of the spectre of Metricom, whose Ricochet service had a rhapsodically loyal wireless Internet access subscriber base that somehow never managed to grow enough to get Metricom to profitability, despite years of marketing and an unlimited usage model. And the Ricochet service didn't limit users to the "Wireless Internet" ghetto, either. They got the full ride.
And, despite the clear lesson of the Internet that users are strongly resistant to the notion of paying for content, the 3G geniuses seem to be convinced that they'll be able to charge admission fees to their subscribers who access Internet-based content from their cell phones. Still less sensibly, they seem convinced that they can persuade those same users to pay only slightly lower fees to access "Internet-like" content that they themselves will own and control -- conveniently overlooking the fact that the cellular industry has collectively zero experience and/or expertise in building Internet portals of any kind, much less compelling ones.
If I were a betting man, I'd wager a good two bits that more than one of these wild-eyed optimists will soon wind up in a big, smoking, bankrupt crater -- and that it won't be all that much longer before enough of their peers join them to turn the European 3G vista into a moonscape.
Worse still, while the entire 3G party depends for its success on the magic of economies of scale, it looks as though Hell will be exporting icewater before the USA -- the single largest market in the world -- joins the festivities.
I, Me, Mine
Over here, the problem of spectrum license fees has already raised its ugly head in the form of wannabe PCS carrier NextWave. NextWave overbid to the tune of more than $4 billion for its 2G licenses, then took refuge in bankruptcy in order legally to renege on payment for them. After the Supreme Court overturned the FCC's repossession of those licenses, NextWave recently let itself be bought off by winners of the FCC's re-auction -- who, between them, paid north of $17 billion for that muffler.
But, where 3G is concerned, innovative, new variations on that particular grift will have to hang fire until the poohbahs at the FCC and NTIA manage to resolve the question of exactly where in the spectrum domestic service will reside. And, like a mirage, the more closely the responsible Federal agencies approach that resolution, the more swiftly it seems to recede.
The rat in this particular woodpile is that, in order to be fully interoperable with European and Asian 3G offerings, the domestic version must inhabit either the 1710-1855 MHz or 2500-2690 MHz bands. Unfortunately for the eager carriers, those frequencies are already inhabited -- and the existing tenants, understandably enough, are more than a little reluctant to be evicted from what they've come to think of as "their" property.
While the inhabitants of the 2500-2690 band -- namely Multipoint Distribution Service (MDS, also known as "wireless cable TV") and Instructional Television Fixed Service (ITFS) operators -- might be amenable to relocation, (asssuming the 3G folks are willing to pay their estimated $70 billion moving expenses, of course,) the current residents of the 1710-1855 MHz estate include the U.S. military. And, given the events of September 11, 2001 and the resulting "war on terrorism", there's precious little reason to think the armed services are going to be willing to put up with the inconvenience, downtime and general distraction of shifting their training traffic and satellite control systems to new frequencies any time in the near future, merely to accomodate a civilian telecom technology of unproven merit.
Which is why the FCC is now considering various allocation options that would essentially kill off the whole fantasy of global 3G roaming altogether.
Bring it on Home
None of this is news to the 3G industry. Almost a year ago, Keiichi Enoki, head of DoCoMo Japan's iMode 2G Internet service, admitted that the vast majority of 3G traffic is likely to consist of voice communications, just as was the case with its predecessors. And, while such mobile luminaries as Thomas Dolby Robertson of Beatnik, Inc. are convinced that teenagers will adopt 3G technology to exchange audio clips of favorite pop songs and the multimedia equivalent of collectible trading cards, the looming global economic crash augurs otherwise.
Meanwhile, it looks to be a terrible economic winter for Mariposa businesses. And -- what with tumbling international currencies, dire predictions of an extended planetary war against terrorism and the resulting precipitous drop in air travel of all sorts -- the prospects for next year's tourist season are equally grim. I fear for the survival of this year's crop of new business entries and, yes, even for some of the established concerns.
Still, if you're the kind of cockeyed optimist who regularly draws to an inside straight, there's always the possibility that the Cassandras are full of what comes out the south end of a northbound bull. Reports of the demise of both Mariposa's economy and 3G alike may, in fact, be greatly exaggerated.
But I know which way I'd bet on both outcomes.
Rod Ghani's comprehensive overview of 3G technology
Alastair Angwin's summary of wireless technolgies and directions
IBM's Pervasive Computing Glossary (handy to parse the alphabet soup of acronyms)
The Wireless Multimedia Forum's relentlessly upbeat Business Case White Paper (in PDF format)
Jupiter Media Metrix's gloomy assessment of 3G multimedia demand
The Yankee Group's similar view of European GPRS overpricing
Deborah Durham-Vichr's carefully-nuanced critique of 3G business models and application requirements
FCC Final Report on 2500-2690 MHz spectrum sharing (in PDF format)
Appendices to the FCC Final Report on spectrum sharing (in PDF format):
NTIA's companion report on 1710-1855 MHz spectrum reallocation (in PDF format)
. . .
A somewhat different version of this work was first published by IBM developerWorks at
(Copyright© 2001 by Thom Stark--all rights reserved)