The chief marketing office of Ericsson, a handset maker, says that Wi-Fi hotspots will be increasingly irrelevant: This story has some legs because it's so outrageous. But let's examine what John Bergendahl means.
From a handset perspective, the increasing availability of 3G, its ever-faster speeds, the roadmap for 3G's evolution and 4G services, the capabilities of handsets, and the services that people actually want on handsets (viewing movies, streaming video from YouTube, taking and sending high-quality photos) are all factors that make Wi-Fi less relevant.
In Europe, Asia, and America, there's enough capacity and enough advanced devices to do interesting things now, but usage hasn't grown fast enough--partly due to excessive pricing--to drive aggregate speeds down for users except in the most congested areas. I've heard scattered reports of people seeing 3G slowdowns at conferences and so forth. The 2.5G EDGE network basically failed at Macworld Expo last January because of the thousands of iPhones all trying to grab a slice of limited spectrum in San Francisco.
Bergendahl sees the challenges as coverage, availability, and price. That's all true, and in Europe more so than in the U.S. Europe has better coverage and availability, but the price for roaming outside one's home country or network is extraordinarily high. Some voluntary efforts to drop roaming prices are underway to forestall 3G data price regulation by the European Commission, such as went into effect 25-June-2007 for voice roaming.
The problem is that he is thinking as a handset maker: he's thinking about capabilities, selling more handsets, and overall revenue from value-added services that he can make sure his devices deliver. This is fine. But it's not how carriers think. There's a growing disconnect between capabilities built into handsets and those offered by carriers. Nokia's insistence on building somewhat open-platform phones with Wi-Fi and video capabilities have hardly been leapt on by European carriers, and those devices aren't sold at all in the U.S.
Really, Wi-Fi is a heat-sink, a complement to 3G. It's a way to inject bandwidth into a network at fixed locations where someone might sit to watch a video or carry out some task that involves being static. You can make phone calls in motion, but you're rarely jogging or driving while watching a video or composing email. (Okay, studies show lots of emails written by drivers. Still.)
Wi-Fi can be fed through direct wired network connections, allowing carriers to offload bandwidth-intensive tasks without disallowing them. Apple, for instance, only allows its iTunes Store to be used over Wi-Fi on an iPhone or iPod touch--as the iTunes Wi-Fi Music Store--rather than stress the EDGE network's low capacity.
You can see how T-Mobile and BT are pairing Wi-Fi and voice, and building networks that allow them to compete for the best cellular customers, letting those customers talk longer but use a much cheaper medium over which calls are placed. AT&T hasn't gotten the religion yet about pairing Wi-Fi and cellular plans, but that's clearly coming, and with a 17,000-plus hotspot U.S. market, we're going to see some new ideas from them, too.
Really, 3G doesn't compete against Wi-Fi because the same operators that run 3G networks can benefit directly from Wi-Fi networks. Until 4G networks are built, Wi-Fi's local network speed and its typical backhaul speed will far outpace what cellular can deliver, and occupying cellular frequencies with big downloads is a poor use of scarce frequency over which other revenue can be better extracted.