This is part 1 of a series of my lay opinions on proactively maintaining your corner of the web.

I recall when linkrot was the major issue for the web back in the late nineties. While it remains a problem today, I think that the web as a sandbox of emergent technologies has not only promoted linkrot, but something even worse. For now, let’s call it service rot. Today, we take for granted site services such as Google Analytics and Domains, URL shorterners like Bitly and TinyURL, and even social networks like Facebook and Twitter. Has it occured to anyone that these services will one day be gone? Is anyone else worried about how dependent we are on major services to promote ourselves and our brands?

I worry because back when I started blogging, oh, six or seven years ago, I tested the waters with Blogger. I had no comments, no blogroll except for the links I would manually edit into my template. Haloscan gave Blogger comments before it was bought by Google. Blogrolling gave me a link manager before anyone else could. Later on, I discovered B2, found a remarkable web host, and my site’s exposure grew just as the blogosphere did. I was an avid participant in the Conservative area of the blogsophere, and all sorts of little kaboodles came out. I made it to a large mammal in the TTLB ecosystem. I played in Blogshares. There are things I did that I don’t even remember doing.

As the number of participants grew, the fidelity of the services floundered. Haloscan would time out. Blogrolling would time out. TTLB? Time out. Blogshares? Well, people just stopped playing that. I can’t document any and all of the services that have gone through this cycle, but the general drift is that as these services grew, as the freeloading users refused to pay, and as the search for funding became more and more frustrating, some services just threw in the towel.

Epochs in the growth of the web are ushered in by groundbreaking ideas that attract capital in all its forms in the hopes to monetize these ideas. So many magazines and websites devote so much time performing case studies on the successes. What few people ever talk about is the disparity between the failures and the successes. Case studies of the failures would probably serve better as cautionary tales for those who have a great idea but do not know how to run a business, because the successful companies out there on the web are where they are not just because of superior product and marketing, but because they are aware of the mistakes their competitors made. They are able to anticipate problems that are a mere matter of eventuality.

The most obvious factor is the economics of scale. It affects us in business “in the real world,” and it affects blogs, online stores, and brochureware sites. It affects all of us, from Google’s massive array of free services to the lowliest blogger who pays five dollars a month in hosting. Each site is by nature a service that is provided to the rest of the web. As the number of service consumers grows, the operating costs for each grow as well. I conjecture that this growth is not so much linear compared to the growth of customers, but exponential, with a very small value for k. In non-mathematical terms, for most of us on shared hosting, or who are doing this blogging thing, or this plugin development thing, for free, or to build enough exposure to get some kind of money elsewhere, there is always a tipping point where the costs go beyond what we can afford, and once that point is reached, there is very little time before we have to shut things down all together.

I have seen this in the political blogosphere, where folks with great opinions and good research get pushed onto the stage, front and center, as when a post “goes viral.” Most of these authors don’t know their ways around anything beyond simple FTP, and are unprepared for the fabled Instalanche. Glenn Reynolds himself has complained, at the height of the MT vs. WP firefights, of WP-powered sites he’d link to that are unable to withstand the traffic. I’ve known a few people who’ve had to switch hosts multiple times with very little time in between moves as a result of CPU cycle abuse. The process is far from painless, and while paying someone to do it can ease the situation, it’s nonetheless a hassle.

As services grow and fail like the ebb and flow of the tides, how are we to cope? My suggestions in Part 2.