I came across two different articles / blog posts today that discussed the decline but also the staying power of legacy services. The first is a blog entry by David Pogue of the New York Times about dial-up. Here’s an excerpt:

From the mailbag:

Dear Mr. Pogue: Can you explain why big sites like Adobe and Microsoft list download times for 56k modems?

I can’t image that 56k modems are used by the majority of U.S. Internet users (or even in the Western world, really). Wouldn’t it make more sense to tell us what the download time would be for a DSL or cable connection (various speeds)?

I believe my reader is correct: dial-up modems now represent well under half of U.S. Internet users.

Do Web sites list dial-up download times on the premise that dial-up users are the only ones who care how long it will take?

My guess in answer to this question, incidentally, is that probably no-one is responsible for reviewing that policy from time to time to see if it still makes sense. It’s probably apathy as much as anything else that’s keeping those download times in place.

The second is a Wall Street Journal column about landlines. Again, an excerpt:

In last week’s Real Time column about cellphones, I wrote that “a call on my landline is almost certainly a wrong number, a charity or a recording of a politician.” Which led one reader to ask a very reasonable question: Why keep the landline?

The answer, unfortunately, involves throwing my wife under the bus: She’s the one who wants to keep it. But her reasons are pretty good: She notes that our cellphones aren’t charged “half the time,” which I’d dispute specifically but not generally; we have a good phone number that’s easy to remember and to dial; and we’re listed in the directory, without which there’d be no way for people who’ve lost track of us to find us. When I said people would just Google us, she said “I’m an old fogey, what do you want?” which is a kinder version of “Shut up, dear.”

Let the record show that my wife is in no way an old fogey. But the record also shows that the tide — at least in the U.S. — is running against her. A report released last week (see the PDF here) by the Center for Disease Control and Prevention paints a picture of a rapidly vanishing landline business. In the last six months of 2007, at least 15.8% of U.S. households had at least one wireless phone but no landline, and 14.5% of adults — 26 million — lived in a household with only wireless phones. (The CDC’s first such survey, conducted in the first half of 2003, found 2.9% of adults living in wireless-only households.)Throw in the 2.2% of households that had neither wireless or landline phones and 18% of U.S. households are without the traditional phone service that was part of our common culture for generations. Landline phone penetration is now what it was in the early 1960s.

The point is that both of these articles are about technologies in decline. In the case of dial-up, of course, it’s a much newer technology that is nonetheless much further along in the decline - only a small minority still have dial-up Internet connections, whereas landline owners are still by far in the majority. But in both cases, those technologies are going to reach the point where networks and services are being preserved for a smaller and smaller number of customers. At some point, the providers of those services will have to flip the switch on those services to “off”.

They’re not the only such services, and it’s a tricky thing to grapple with. We’ve recently seen the switch-off of AT&T’s TDMA network, and in early 2009 we’ll see the shutoff of the analog TV network. Sprint has caused some problems for itself by providing an off date for its legacy business networks, giving its competition some easy ammunition. At some point, all companies will reach the point where the cost of maintaining these networks for the laggards is outweighed by the benefits of forcing a switch, but the trick is always calling that point accurately. And the downside is getting it wrong and causing a huge customer service and/or sales problem. Luckily, landlines are going to be around for quite some time, but that dial-up market is going to be heading for the chop rather sooner. When does AOL flip the switch on that?