We all want to be number 1 on Google. For most people that is the only target, but what does that actually mean in terms of results, and how do the different ranks affect traffic to your website?
There have been a number of research experiments conducted by universities over the last decade testing user behaviour on a number of search engines with many using methods unique from the others. However, one technique has been more successful than others – eye-tracking.
Simply put, eye-tracking is a type of software that records and measures eye movements. This method has been used in several experiments for measuring attention given to each site in a search engine results page.
Of these tests, only a few tracked the searchers’ click rates for each position, and of these only one was conducted using Google (‘Eye-Tracking Analysis of User Behavior in WWW Search’). And let’s face it; Google is the most popular english-language search engine around. It has even become a part of our language (have you ‘Googled’ something before?)
In the above experiment, the researchers at Cornell University used a total of 397 queries and recorded the average attention time and number of clicks given to each of the 10 search results on the first page of each query, and the first result on the second page.
Only the initial clicks were recorded (i.e. this did not measure the number of users that returned to the search results to click on another website), and the queries were cached to a local server to ensure that all users received the exact same search engine rankings. Fig A shows the recorded statistical data, and Fig B shows the results in a graph.
|Fig B – Clicks vs Attention for each position in the search results|
One additional statistic that the above table does not mention, but that some of you may find useful, is that for the searchers that clicked on one of the search results it took an average of 7.78 seconds to make a selection.
Here is a summary of the most important deductions from the results of the experiment:
A number of theories can be reasonably concluded from these findings. Firstly, the number of searches done each month for a given keyword does not indicate the number of people that will click on a search result – many (approximately 32% according to the research) will refine the search to be more relevant to their own needs.
The differences between positions one and two were the biggest puzzle to me. However, after considering the statistics for a short time I realised that when a query appears most users, including myself, will initially read the top search engine ranking assuming that it is the most relevant website, then read the second position to compare to the first and see which is better. Then, the user will return to click on the first position for one of two reasons: either the second position was not as relevant as the first, or it was very similar and clicking on the top position feels more reliable simply because it is ranked higher by Google. There would obviously be other influencing factors in the statistical anomaly between number 1 and number 2, but this seems the most plausible explanation.
Once users begin scrolling they only skim the search results, as opposed to the top 5 which will appear on the initial query screen automatically. Many users may read the first few positions in slightly more detail, scan down the page briefly to see if anything stands out, and then return to the top of the page to click on the higher ranked websites. After some analysis of my own search techniques and that of my friends and colleagues, I found this to be accurate in the majority of cases.
Being on the second page of search results will only ever be significantly beneficial if the keyword in question has extremely high search volume. If there are 100,000 searches per month for a single exact keyword phrase then ranking 11 will still bring in 2,000 visits for the month for that keyword alone. But if there are only 1,000 searches and you are ranked 11 that number drops to only 10 visits for the month. So if there are only a small number searches for a given keyword then you will definitely want to be in the top 5 for that keyword phrase to see any significant results. On the upside, if there are only a small number of searches in a month then, in most cases, the phrase will not be hugely competitive.
To build a brand name online, clicks may not be as important as attention. As long as people are seeing your website or business name on a regular basis, your brand awareness is growing. This would be ideal for targeting area-based services like transport, plumbing, electricians, lock smiths, etc, and would be more effective if the service name is in the title or the URL. For example, newyorktaxis.com, manchesterelectricians.co.uk, or sydneybuses.com.au. In cases like these, being number 1 for a given keyword may not be as important as being in the top 10 for 100 related keyword phrases. The idea is that if you can get your name noticed as frequently as possible then when the time comes for the user to require your services, the attention given to your company name can attract the user to your particular website, even if it is not in the top position. You could call this “online branding”.
Now if we get back to the original question, how do search engine rankings affect website traffic? Fig C shows the implied percentage of traffic received by each position in Google’s search results for a given keyword.
|Fig C – Percentage of total clicks received by each position in the search results.|
There will always be other factors influencing user behaviour but, looking at the chart, it is reasonable to say that being number 1 has its benefits, wouldn’t you agree? Of course I don’t think the researchers took Wikipedia into account…