If you search Google for PowerBook 1400, wireless PowerBook, or even Wi-Fi 802.11b, an article I posted here last year comes up within the first few results—sometimes even above Apple's pages on a similar topic. How did that happen?
This isn't one of Google's imperfections. If you have an older PowerBook and you want to put it on a wireless network, or even if you just want to learn about Wi-Fi networking, my page is probably pretty useful to you. What the results do show is that writing for Google works. But we're not really writing for Google; we're writing for other people on the Web.
I wrote the article because I had trouble finding accurate and complete information online about getting an old PowerBook onto a wireless network, even though I knew it could be done. It took a lot of trial and error over several weeks to get it working, so I wrote up what I did to save other people the work.
But that was no good unless people could find the page. So, knowing a bit about search engines in general, and Google in particular, I did a few things:
- I put it on its own page, rather than in with a bunch of my other journal entries, so the page itself would be entirely about the one subject, and not a bunch of different things, like the page you're reading now.
- I gave it a long and almost excessively detailed title, which you see in your browser title bar and maybe in your bookmark field if you bookmark the page. In the HTML code, it looks like this:
<title> AirPort | Wi-Fi | 802.11b wireless networking for the PowerBook 5300/1400/2400/3400/old G3/190/520/540/500 series - Derek K. Miller </title>
- The first heading on the page is also a monster, and quite similar: "Wirelessly Networking a PowerBook 1400 or Other Old Apple Laptop: Step By Step, Covering AirPort, Wi-Fi, 802.11b, and WLAN with the PowerBook 5300, 1400, 2400, 3400, 500 Series (520 and 540), 190, and G3 with Mac OS 7.5 through 9.2." So I packed pretty much any term I could think people might be searching for into the title and level 1 heading, because I knew Google and other search sites put a lot of stock in them.
- I made sure Google, Yahoo!, and the Open Directory knew about the page, and I also pinged Weblogs.com about it, since it is effectively an extension of my weblog page. (If I were doing it today, I'd use Ping-o-Matic instead.)
- I posted about the article in a few relevant places, and mentioned it to the people who run some relevant sites. Since my article was pretty good, they noticed it and linked to it, making Google take further notice. Note that I wasn't just randomly spamming news and information sites—I only posted in places where my page might be genuinely interesting.
There's nothing nefarious in any of this. What it comes down to is, (a) I wrote the best article I could, and (b) I let people on the Web know about it, and in letting people know I tweaked the article a bit so it included most of the words they might be looking for. If the article sucked, it wouldn't get linked to, and it wouldn't show up high in Google rankings for any search in particular, no matter what other effort I put into it.
So I wasn't writing for a machine. The key is that people make the decisions to link, so ultimately, for a page to come back high in a list of search results, people need to like it. That was Google's great insight in 1998—and now, when those who make web pages make links, we're doing Google's work for it.