The Atlantic Gets SEO-Ganked by The Huffington Post: A Breakdown

Last week I was reading some online commentary about a piece by Matt Taibbi in Rolling Stone, this one about Goldman Sachs. I first read Taibbi in Spanking the Donkey, and I usually like the dude’s cynical, New-Journalism-style writing, even when it’s a bit sensational.

The Goldman Sachs thing — at least the parts I could read, since the whole thing isn’t available online (very Time Inc. circa 2003 move by Rolling Stone) — so then I started reading some of the criticism. One of the articles I read mentioned that Megan McArdle (who I haven’t read before) from the Atlantic wrote a big response to the article, so I decided to go try and find that. I typed this into Google:

atlantic goldman sachs taibbi

But I ended up with this order of results:

Snap, it looks like this Huffington Post piece is eating this Atlantic piece‘s lunch, even when I’m looking specifically to go to the Atlantic page.

We all know search-engine listings are important, so how did HuffPost pull this one off? When I look at the source code, it seems like The Atlantic is missing some simple SEO best-practice stuff that could have helped them out here. HuffPost isn’t kicking ass at SEO coding, exactly, but they’re doing enough of the small stuff right that it’s likely that was enough to get Google to list them over The Atlantic. Running down the basic-level SEO checklist:

• Friendly URLs: Sort of a wash; both URLs contain the keyword strings “taibbi” and “atlantic” at some point.
• Header tags: Running into some trouble here — HuffPost has the page title “The Atlantic: Taibbi Is ‘Becoming The Sarah Palin Of Journalism'” in an h1 tag, while “Matt Taibbi Gets His Sarah Palin On” is an h3 tag on The Atlantic. That’s a weird one, particularly because there’s no other competing h1 tag on the Atlantic page and we’re looking at an individual post here, not the main page of the blog where you might use the h1 for the blog title.
• Meta keywords: Keywords are one of the few meta tags that actually count for search engines, and while they’re downplayed a lot in best-practice techniques compared to the days when people would put 200 keywords into the source code, they still serve a contextual purpose. The Atlantic didn’t even use any — not good, not even as good as the cursory list on the HuffPost page.

Plenty of SEO-consultant advice is voodoo, as Google’s algorithm is a mystery to most, but low-level code tactics like URL structure, meta and header tags make a difference.

HuffPost gets some flak now and then for not doing enough original stuff (or sometimes going beyond that), but in cases like this, they’re grabbing the traffic in a perfectly legal way. Even if they do link through in the end, this is a tangible example to pay attention to best-practice coding when you’re building a site, because otherwise, web dudes will come and G up even those readers specifically looking for your stuff.

Webdev word is bond.