There is no relationship between the position of an advertisement in the Google Ad listings and the chance of that ad being clicked on. Bidding more per visitor in order to get a higher position will not get you more visitors. The number one position in the listings is not the best position. No ad position is any better than any other. The factor which has the most bearing on your chance of being clicked on is the text in your ad, not the ad's position.
These are the conclusions I have come to after analyzing the Google Ads I have been running this year. This article will show that, while position in the listings used to be important, it is not anymore. People are more discriminating in their use of Google Ads than they used to be; they have learned to read the ads rather than just click the first one they see.
I have run Google Ads on behalf of clients and for myself for over four years. I launched my first Google Ad campaign in April 2002. Recently I conducted an analysis of the ads I had run in October and November of this year in order to determine the best position for some new Christmas ads. Some people claim the number three position, at the top of the right-hand column is best, others claim the number one spot, so I decided to look for myself. I was surprised, and this article grew out of the investigations my findings prompted. I don't pretend this is a definitive analysis, but it has convinced me things have changed in Google Ads and there is no longer any need to bid for the highest position. This is potentially a very important development, so I urge you to undertake this analysis on your own ads for yourself.
The following chart shows the average clickthrough rate for each position on the first Google results page for a number of different campaigns, in different industry sectors, over October and November 2006. The clickthrough rate is the vertical (or Y) axis, while the average position is the horizontal (or X) axis. Google's "average positions" are not whole numbers. If an ad's position is shifting equally between one and two, it will have an average position of 1.5. There are therefore 100 "positions" on the first page of the Google Ad results. I did not examine positions beyond 10, which are on the second and subsequent results pages. While there is a surprisingly high chance of being clicked on in these pages, no one I know really wants to be listed there if they can avoid it.
Here is a chart of my results:
As you can see, there is no real prominence to the top positions-- people are clicking ads equally throughout the listings. If the top positions were more likely to be clicked on, the chart would be higher on the left-hand side, and taper off to the right. The only positions which stand out as more likely to be clicked are those at the bottom of the page.
These results surprised me. I was sure that at some stage in the past higher positions had generated more traffic. I therefore decided to run a year-by-year analysis for all the campaigns it was possible to get Google Reports for. Were we ever correct in thinking position mattered? If so, when did things change?
My analysis included around five million ad impressions which had generated around 40,000 clicks over four years. While these are not astronomical numbers for Google Ads, they are sufficient to ensure the results are statistically valid. The campaigns were run mainly in the UK, but also in the U.S. and France. The ads in Google France were written in French (Tip: Don't place English-language ads in Google France unless you want to irritate potential customers). The sectors covered were accountancy, mortgages, real estate, tourist attractions, car insurance, restaurants, hotels, software, and office services. Clickthrough rates overall averaged 2.3 percent. The lowest click-through rate was 0 percent and the highest was 100 percent. The average price-per-click paid was 56 cents. The lowest price-per-click paid was five cents and the highest was $19. By covering such a disparate set of sectors spread over three countries and two languages, together with a wide variation in success and bid rates, I eliminate variations based on consumer profile. I used only ads which ran in Google itself, not in any syndicated outlets, so there are no complications associated with variations in display.
My analysis involved taking the average position and corresponding click-through rate for each keyword each month, a total of around 1,500 pairs. I then calculated the average of all the clickthrough rates for each position across all keywords.
These results cover a long period of time, almost since search engines first started offering PPC ads on their results pages. In that time the profile of people using the web has changed, and their search behavior has become more sophisticated and discriminating. By analyzing the results on a year-by-year basis I hoped to understand if, and how, people's click-through behavior had changed. Even though I had first run ads in 2002, I didn't have sufficient data for that year to be satisfied I was guaranteed a statistically valid population, so I commenced the analysis at 2003.
As can be seen from the chart, there is some tendency to click the visually prominent positions at the top. In 2003 Google Ads were confined to the right-hand column of the page, so these were the top two positions in that listing. However, the spike for positions five to six shows that being in the top position is not the only place which earns a high clickthrough rate. On most screens at the time positions five to six appeared in the bottom-right corner of the screen, so perhaps this accounts for the spike.
Next: Trends in 2004 through 2006