When I suggested last month that someday, search engines could work without links, a lot of people thought it was an interesting gedankenexperiment (they didn’t use that word but hey, it’s one of my favorite words) but didn’t believe Google is really moving in that direction. The future, by definition, remains to be seen, but today in his webmaster video, Matt Cutts revealed that Google has already experimented with a version of Google that doesn’t use links as a signal:

A user asked:

Does the big G have a version of the search engine that totally excludes any backlink relevancy? I’m wondering what search would look like and am curious to try it out.

And here’s Matt Cutts’ response (emphases mine):

We don’t have a version like that that is exposed to the public, but we have run experiments like that internally, and the quality looks much much worse. It turns out that backlinks, even though there’s some noise and certainly a lot of spam, for the most part are still a really really big win in terms of quality of search results. So we’ve played around with the idea of turning off backlink relevance and at least for now backlink relevance still really helps in terms of making sure that we return the best, most relevant, most topical search results.

As I said in my post on why guest blogging isn’t dead, it seems pretty obvious that Google’s algorithm is still hugely dependent on links. Further, Google can’t algorithmically tell the difference between a contributed article, a post on a multi-author blog, and a guest post if all the other quality signals are the same, so they’re trying to scare spammers into backing off a scalable tactic. But the fact is those links still help.

Here’s what I find interesting though:

  • Google recognizes that links are highly spammable, and if they knew how to filter out all the spam they wouldn’t have to scold and spook us so much. They’d just take care of it. (Email providers don’t spend their time telling marketers what not to put in emails – they just filter out the spam.) Google may be outpacing spammers for now, but spammers are catching up.
  • Google recognizes that there are other potential approaches to ranking results. They’ve already tried some out! Cutts doesn’t say what those internal experiments were based on, but we’re guessing the “Quality Score” style of algorithm as laid out here was where they started, looking at engagement metrics like expected click-through rate based on position.
  • Google doesn’t yet have a way to rank search results that is more effective than the link graph – but they have people working on it, and either way the door is open for other companies to find new ways to map the web.

I think Cutts’ phrasing is telling, especially that “at least for now” and “still really helps” – the implication is that they want to move away from links or at least reduce their dependence on links as a ranking factor, but just haven’t formalized a new system yet.

What do you think? I know I’m not the only one who would love to see that experiment!