Mar 25, 2012

How Webcrawlers Make the Web Worse

Reading about how to improve small business search rankings reveals the way webcrawler design leads businesses to create so-called "content" that does everyone a disservice --- both their own customers as well as people actually looking for useful information.

I started reading about "search engine optimization", or SEO, partly at the behest of my dad who's trying to improve his sign shop's website, but also to help with promoting the stuff I do. Anyway, SEO tends to be a lot of fairly boring technical little stuff, mainly about all the relentless ways you can squeeze in "keywords" and "keyphrases" into every last pixel of your site, for the benefit of Google's (and others') webcrawlers and "robots". But then it got interesting (or perhaps it only seemed so by comparison) when my RepKoverTM bound guidebook (lays flat for easy reading) started advising me to simulate "content" to draw in the flies --- to get some Content Management Software (CMS) and perhaps hire a copywriter to fob off some faux-info on the hoi polloi (assonant aside: I dream of one day opening a competitor to Hobby Lobby, either catty-corner or across the street, to be called The Hobber Lobber. Additional aside, non-assonant: perhaps these very words are only a content simulacrum, cynically generated to ensnare you in some commercial agenda?)

I do have a point here though so bear with me. It's not all alliteration, assonance, and arcane lingo. I'm just easily distracted like that. O.K., so the SEO experts are advising businesses basically to pretend to be "informational" sites, in order to draw in links to their so-called "content" and generally improve their appeal to webcrawlers, which favor that sort of thing. A webcrawler is like a little old lady (or me) it would seem, it just wants to curl up with a cup of tea and a cat on its lap and a good magazine. And some hierarchical content. Hierarchical content is key. H1 ("Heading One"), H2, H3, and so on. "P" is for "paragraph". ("TT" equals "teletype", in case you're ever coding in HTML and need some teletype.)

But it seems to me like this is all wrong-headed on the part of the 'bots. Sometimes, to put it simply, all a user wants is to find a local small business --- not an "information-rich web resource" about sign production, or different facets of the sign industry. Not a &%$# blog about the latest tweaks to vinyl formulations (signs are made out of vinyl nowadays, ICYDNK.)

All I'm saying is, Why CAN'T someone just have a website that's like a slightly expanded phonebook display ad, basically? And still get good rankings? A few pictures, a list of services, a fetching motto, contact info, and you should be good to go.


BZZZZ. Wrong. Bots hate that. Bots would prefer you grab some generic fair use material about whatever business you're in, maybe change a few words so it doesn't get down-graded as duplicate content (check with your own SEO expert about this one, I'm not quite sure how it works), and maybe also turn the "keywords" into links. Another good plan, in the view of article-craving bots: outsource your copyrighting to someone in India (I'm going to suggest this to my dad) who can come up with at least two or three topical articles inbetween doing some online shopping for you and making sure your kid's homework assignment gets sent to your Blackberry for you to review (or so I hear, I don't have any of that stuff. And by "stuff" I mean "kids" or "Blackberrys".)

So, we non-bots (if you are a bot, just skip this part) get stuck with a lot of content online that's coldly calculated to appeal to bots, which is really just businesses pulling said "content" out of thin air that they have no actual interest in writing about, other than so far as it can improve their rankings. Instead of simply building a functional site that makes sense for them and their users. Maybe you've come across some of this uninspired prose while you WERE actually looking for info. It tends to read like it was translated from a foreign language.

It would be interesting to know the reasons for this bias on the part of the bots --- whether the designers were exercising a value judgment and trying to exclude the crassly commercial, or whether it was simply easier and more intuitive to program them in a way that favors "informational" sites. Regardless of why this is, by addressing their rankings to only one kind of user (the tea-sipping, New Yorker-magazine-reading type) they force America's proud business-men to game the system, to simulate content, to outsource jobs (hey, American copywriters are expensive!) And so we must fight back. We must come up with webcrawlers that privilege information-poor, no-nonsense, straight-to-the-point websites. Not long, rambling blog posts that go on and COMICS on and on and CARTOONIST laced with keywords that just keep HOT COMICS going on and MUST BUY COMICS on...


  1. Very nice indeed! Thank you for the guide about web crawlers.

  2. I don't really agree with the assessment, but tell me this, Mister Rocks - when was the last time that you tried to teach a machine to think?

    The bots are playing catch-up. It will be a while before they get to your lingo jingles, yet.