In 2000, watch Amazon.com go the way of Netscape.

As much as I love Microsoft, they really did gang-rape Netscape. Netscape's browser was and is superior in just about every way to Microsoft's Internet Explorer. But you can't really fault a company in giving something away, even if it is a slightly inferior product. Unlike Netscape, however, Amazon.com is strategizing itself against a future that is approaching more rapidly than anticipated. CEO Jeff Bezos' greatest success to date may be his continuing ability to convince his shareholders that, in the short term, profitability and strategic growth are mutually exclusive. The problem is that his short term doesn't seem to have a deadline, and you can only maintain a market capitalization of more than US$25 billion for so long without making a profit.

Which isn't to say that Amazon.com won't figure prominently as a high-profile commerce engine. Between the company's zShops (bringing Mom and Pop e-tailing to the Amazon tent) and its expanded product offerings (adding software, hardware and home improvement items, among others to its inventory of books and CDs), Amazon.com enters the year 2000 as the unchallenged leader among online stores. But these new forays aren't cheap. Throw in the seasonal advertising blitzkriegs and you've got a money-losing company going further into the hole as niche competitors arrive on nearly every front. Consider selected fundamentals:

• revenue per customer is slipping (US$32 in 2Q98 to US$26 in 2Q99);

• customer growth rate has fallen off (currently 5 percent, compared with 36 percent a year ago);

• customer acquisition costs increased nearly 50% from 1Q99 to 2Q99.

Netscape's IPO was certainly a benchmark. The e-commerce numbers from the 1998 holiday buying season were another. But through all of this rapid change, one thing has remained constant: The lowest common denominator for accessing the Internet continues to be the annoyingly slow and cantankerous 56K modem.

Yes, many of us scream through the Web via the office network with its high-speed Internet connection. But among consumer, or residential, users, the modem is still the primary option. And every year (we're going on about four years of this now), we hear about how great the Internet will be once we all have broadband, or high-speed, connections to handle all kinds of new "content," from video to music to games.

The good news is that high-speed options are gradually filtering down to end users, primarily in the form of cable modems and digital subscriber lines (DSL, for short). To recap, cable modems are based on the cable television network, modified to handle both inbound and outbound network traffic. DSL operates over the existing telephone grid of copper phone lines. Both technologies have their limitations. Cable modem service tends to degrade in performance as more users join the network. DSL can be a bitch to install, costs a bit more, and its service area is limited to within a mile or two from a telco central office.

By the end of 1998, there were only 50,000 DSL customers nationwide. Through 1999, that number has since jumped to more than 600,000, with hot growth projections as more access equipment is installed in telco central offices around the country. As of 2000, it is estimated that 80 million homes will be able to use DSL.

At the same time, high-speed Internet access is also moving quickly to cable. The primary retailers of Internet access through cable modems, Excite@Home (ATHM-NASDAQ) and Road Runner, ended 1998 with 500,000 subscribers. Now, they have 1.3 million. But as of 2000, both companies will have access to 68 million homes, 15 percent fewer than DSL.

The geek debate over which technology will win is hot enough to cause a riot at the next Star Trek convention. Watch DSL nose ahead of cable through the final turn in 2000 for two reasons. First, DSL is strongly positioned to serve the small- to mid-sized business market.

DSL holds a distinct price advantage over leased line alternatives here and cable modems are simply not being effectively marketed to this customer segment. Second, the DSL wholesale companies--Covad Communications (COVD-NASDAQ), Northpoint Communications (NPNT-NASDAQ) and Rhythm NetConnections (RTHM-NASDAQ)-- have demonstrated keen savvy in forming the right partnerships. This is key with regard to distribution, as all three companies are selling DSL through hundreds of regional Internet service providers, not to mention national Internet/online service providers such as America Online, Mindspring, and Compuserv. The cable modem companies do not seem to be able to work with anyone they don't own.

None of the wonderful content and commerce implications promised by broadband access will mean a lick until these technologies are fully deployed, which should occur by mid-2001. In the meantime, buy a cable modem while it's available and invest in DSL companies while they're still cheap.

Please read on...