Open
X

Algorithms Matter. But What’s Behind Them?

The idea that Google Search results are ranked using an “algorithm” that measures the relevance and quality of web pages and other pieces of content has become so mainstream that many marketers talk casually about the “Google algo.” It’s even been the subject of a CBC Radio documentary.

A lot of advertisers now, quite justifiably, apply the same type of understanding to the formula Google uses to rank ads. Although increasingly complex, this formula is still basically simple: it’s QS (keyword quality score) X Max Bid (your bid on that keyword) = AdRank. The complexity comes in when you start trying to unravel what Quality Score really is. Back in 2002, Quality Score was one thing: Clickthrough Rate (CTR). Today, Google maintains somewhat opaque help files outlining what other “relevancy and quality” components go into Quality Score. Inevitably, the fever pitch of misleading claims by vendors (many of whom use traditional “SEO style” obfuscation tactics) has trickled out into the marketplace so what should be clear enough gets shrouded in a kind of mystical fog of… yep we’ll say it, self-serving B.S.

Now, more than ever, paid search accounts need to straddle the competing demands of relevance, ROI, and Google’s bottom line. Quality Scores influence your results… heavily. It’s important to understand user intent issues and other trends that could be hindering your account’s overall health.

The funny thing about it, though? Finicky, fine-grained approaches to account management were beneficial even when it was just a “CTR multiplied by your bid” formula, and we laid these out in great detail back in 2005 when Andrew published Winning Results with Google AdWords.

There is a lot more continuity than change inside of Google’s black box algorithm. An unfortunate threat to your bottom line now becomes the tendency to look at the little scores as a “scorecard” and to begin making decisions based on trying to improve those scores. All else being equal, you’d rather have a keyword Quality Score of 10. But the real scorecard remains, as always, strong ROI. And to achieve that is not a one-step process, but a complex process of managing to key objectives over the long term, with user search intent in mind… testing and iterating in relation to a large amount of user behavior data.

That’s where we’ve been living for the past decade. It’s a pretty sturdy structure if you build it right.

NEXT PAGE –>

Technology & Content Partnerships