Five years ago, when a journalist at this paper, we became enchanted by the internet. Netscape, now part of AOL, had gone public, and venture capitalists were pouring cash into companies with brilliant technologies or superb vision, such as Inktomi or Yahoo!
A few million people in Britain were connected to the net, mostly at shared PCs in offices. We marvelled at those with their own web site. Then, I felt bullish that we'd tackle the "world wide wait".
Several technologies would compete to provide high-speed network access, which would become widespread within five years. Five years on, and - with the exception of South Korea - residential broadband access has failed to even start taking off.
Even bullish tech analysts, such as Forrester Research, predict that fewer than one in five British homes will have broadband internet access by 2005. What went wrong? Many, including The Guardian, have catalogued the frustrations and errors on the supply side. Demand-side weakness is as much a contributory factor. Simply put, we have such low broadband penetration because there isn't a compelling reason to pay for a connection.
Broadband offers three advantages over vanilla modem connections:
1) it is faster, typically 512kbps shared across a bunch of other users;
2) it is always on, meaning there is no dialling-up;
3) it is unmetered, meaning most tariffs don't charge for being online.
But for the overwhelming majority, it doesn't allow them to do much they could not do before. Perhaps fraudband would be a better term.
For most of us, 56kbps is pretty good for most needs. Those higher bandwidths are useful for file-sharing, such as Napster, or long downloads. Always-on connectivity is great for running your own web server and online gaming. But these are things that few people bother with. And if they are popular, we haven't yet worked out an economic model for them. Take instant messaging.
It becomes much more valuable with an always-on connection, because you are more likely to use it. Napster is perhaps the best example of a popular broadband application with no economic model. Napster is bandwidth hungry and thrives on persistent connections. But internet service providers (ISPs) worry about its effects, perhaps as much as the music companies do.
Comcast's US cable modem service shut off access to Napster because its network was straining under the load. A large European broadband ISP found that 5% of users were responsible for 85% of the network usage, predominantly exchanging movies, software and music. That will mean another big infrastructure investment that cash-strapped ISPs need to make, if broadband connections are to explode.
Most of what we want to do (email, instant message, shop, browse websites) happens well enough at modem speeds or a little faster. The marginal value of taking a 56k connection to a shared 512kbps connection is small. If all this sounds familiar, it is because we can apply the same lessons to 3G mobile. Most of the benefits of 3G services will be delivered to consumers over the next couple of years with 2.5G upgrades (higher bandwidth than today, always-on connections).
This incremental investment will deliver huge value. By contrast, subsequent 3G upgrades will deliver marginal benefits for a much larger investment. Having been off the mark with my predictions five years ago, I would like a second chance:
1. Analysts will revise forecasts for broadband internet access downwards again, and telcos will talk less and less about their consumer broadband offerings.
2. TV companies, BSkyB in particular, will drive broadband DSL connections with new generations of their set-top boxes to enhance personalised programming and TV shopping. These won't, primarily, be for fast web access.
3. Community-based internet services will spring up using a single DSL or cable modem connection shared with their neighbours on a free-access wireless network. 3Com is already making a low-cost wireless Ethernet device to do this. This will grow the number of always-on, better-than-modem bandwidth homes in the UK. See you in five years to wipe the egg off my face.
• Send comments to Online.feedback@theguardian.com
