Trends in 2004 through 2006
This chart shows that during 2004 people are still being influenced to a major degree by the ad's position. However, the chance of being clicked on is effectively the same for any other position. For example, position three is no better than position six. In other words, if you can't be at the top, it doesn't matter where you are.
A change in behavior is evident by 2005. While there is a considerable spike around the top-right position (between two and three), the number one slot is no better than six (the bottom-right corner on many screens), or the position at the bottom of the entire list. Indeed, by this time many PPC campaign managers had designed bidding strategies to hold the bottom of the first page listings. I think this is the time when some consumers were starting to understand out that not all ads are equal and they need to read them.
By 2006 things had changed considerably. While there is some advantage to the number one and number three slots, all positions in the listings show roughly the same clickthrough rates.
What this tells me is that the days are gone when paying more for an ad's position could be justified by increased traffic. There is no longer any value in paying more than someone else to place your ad above theirs. Being positioned above another ad doesn't mean that you are more likely to be clicked than they are. People don't blindly click the top ads any more, every position has an equal chance of success. I think people now read the ads and decide which one to click on based on the text in the ad itself. I have no evidence to back this up, but it is anecdotally confirmed by people I've spoken to.
What this means is that the text in the ad is more important than the PPC price you're paying. My policy has always been that the ad heading should exactly match the search phrase the ad is placed against. If people are scanning a list of ads, it seems logical to me that any heading which contains the exact words they typed in will get their attention. I have no evidence that this is true, but I'm getting clickthrough rates way above average, so it works for me.
If you follow this logic through, it means you need to run a separate ad for each search phrase. This may seem like a great deal of work for many companies, but it's what I do and I think the results justify the effort. It means you can get a great deal more traffic from your ads, and at a much lower per-visitor cost. It also provides a level of granularity for analysis and improvement which you can't get running a single ad against a large number of phrases. If you're Google Ad spend is considerable, the extra labor building the ads will be more than justified by the lower per-click costs.
You don't have to accept my conclusions. It's easy enough to run your own analysis and see for yourself. However, in my view the world has moved on-- there's no need to bid for top positions any more, any first page position will do. What's important is the words you place in the ad itself. In other words, the winners in Google Ads aren't the ones who throw the most money at PPC, but the ones who exercise the most care and attention to detail.
Brandt Dainow is an independent web analytics consultant. Read full bio.