A website is not a ship

A new crime-mapping website for England and Wales is experiencing a “temporary problem” as millions of people log on every hour, the Home Office has said.——BBC News

A website is not a ship. You don’t crack a bottle of champagne against its bows, push it down the ramp, and wave it off to the high seas. This ought to be obvious. And yet, it seems that many—maybe most—people don’t grasp this fact.

If your website is useful, you don’t need to launch it. It will be useful at any point in its lifespan: today, tomorrow, maybe even next year, insha’allah.

Conversely, if your website is not useful, then—quite apart from the fact that you shouldn’t have bothered in the first place—it won’t be any more useful in the next few hours than at any point in its sad, drawn-out existence.

All that launching your website—with the accompanying spots on breakfast news and columns in the morning freesheets—will achieve is to drive thousands of times the regular traffic to the site in a brief period. Can it handle that? Maybe, if you’ve designed it to handle that. But the time you spent doing that optimisation was almost certainly wasted, because you created the problem that you now need to solve.

I’ve said it many times before, and I’ll say it again: narcissism is one of the biggest causes of technical problems.

A website isn’t a TV programme. It’s not the X Factor, where a significant portion of the nation tunes in to share the ephemeral moment of watching some no-talent assclown gurn their way through a Beyoncé cover. (Caveat: I’ve never actually subjected myself to X Factor.) There is nothing to be gained by having everyone click at once.

If your website is indeed useful, then you certainly do want to get the word out. But a saturation media campaign is not, emphatically not the way to do it, because it will not deliver a good service to the visitor.

There are ways to design a website to handle massive spikes of traffic. The best one of all, though, is to realise that the internet is a pull medium, not a push one, and to work with its limitations and peculiarities rather than pretending it’s just another broadcast channel. It’s not, it doesn’t work like that, and treating it like one will ultimately not work very well.

Make reasonable optimisations. Allow for traffic spikes: if your website is useful, then you should expect those to occur. But don’t try to drive the entire population of a country of 70 million people to your website on day one. That’s just stupid.

Comments

Skip to the comment form

  1. Thomas

    Wrote at 2011-02-08 13:05 UTC using Firefox 3.6.13 on Linux:

    Paul,
    Interesting as always. Was this an intentional play on “The internet is not a Truck!”? Don’t want to clog those tubes on purpose now…

Leave a comment

Please read the comment guidelines before posting. Comments are Gravatar-enabled. Your email address will not be published.

To prove that you’re human, type human in the Bot check field.

Trying to post some program output or a long code sample? Please use a paste service and link to it instead.