The evidence that Google's RankBrain uses click-through-rates (CTR) as a ranking factor means there's now even more reason to improve your organic search listings.
What Influences Click Through Rates (CTRs)?
CTR vs. Rankings
Now Google has removed sidebar ads, the biggest influencer on CTR is simply the ranking of each individual listing. The higher up your listing is shown within the results, the more clicks it will receive on average, which in turn increases its percentage CTR. People trust that search engines will place the most relevant results at the top of the page and many people don’t bother scrolling down the results pages:
Ads appear for most commercial search terms which take many clicks away from the organic listings (between 10% and 40% depending on the keywords targeted):
Takeaway point: We need to consider the ranking of a result in order to check it’s CTR performance fairly. We should expect the CTR to be lower if the position of the listing is lower within the results.
CTR Vs. Brand
Brands are massively influential in people’s decision making online and can heavily influence which search results people click on, despite possibly being lower down the rankings.
When it comes to ecommerce for example, Amazon and eBay are so popular that people will often include their brand names along with their search term. Search terms such as “pc games amazon” or “used smartphone ebay” strongly signify that the user wants to see results from a specific website.
This brand loyalty is prevalent online. People place a lot of trust in websites which they have used before and received a good service from and they are likely to already have an account setup, so the checkout process is much quicker.
Products or services created by particularly brands also influence CTR. For example, there are certain car brands which people always buy, or certain restaurant chains people actively avoid after a bad experience. Each individual will have different feelings about certain brands and this can make products or services have a much higher or lower than expected CTR, even when the brand isn’t actively searched for:
Takeaway point: Branded search terms shouldn’t be considered when looking at improving current CTRs as they can wildly skew the statistics.
CTR vs. Result Page Number
The second page of Google’s search results is rarely visited with less than 5% of people clicking them according to studies done in 2014. The act of clicking on to the next page of results acts as a barrier for some people, unless they are particularly unhappy with the results on the first page.
Mobile users only have the option to view the next page of results (they can’t skip to say page three right away) and typically there are less than nine results per page on smartphones.
Currently, Google has around eight to 10 organic (non-paid) results after a search is performed. This means that click-through behaviour may be far less predictable when looking at data on keywords ranked higher than position eight.
Takeaway point: Looking at CTRs listing higher than position eight may not follow a predictable pattern due to a second page of results.
Drawing Out CTR Data with Search Console
Google’s Search Console (formally know as Webmaster Tools) allows you to download search query data on your top 1,000 keywords. To find this, you’ll need to navigate to [Search Traffic > Search Analytics]:
You’ll want to look at the previous 90 days of data, the most Search Console allows you to see, and enable “Clicks”, “Impressions”, “CTR” and “Position” as you will require this data:
Scroll down to the bottom of the page and “Download” the data so it can be exported and modified in a Spreadsheet program such as Excel or on Google Docs.
Identifying the Best and Worst CTRs
The data downloaded from search console needs some filtering before it can be used to identify the best and worst CTRs.
Bearing in mind factors that unnaturally influence CTRs, you’ll need to delete each row which falls under these conditions:
- Remove any search queries relating to your website’s brand name – Branded search queries always have a very high CTR and will skew the data. For example, on our website HallamInternet.com we would remove any search query with the word “hallam” within it and any similar spellings or typos such as “halam”.
- Remove any search queries after the average position of 8.0 – Queries which appear over position 8.0 on average may not be on the first page of the search results. Results on page two onwards will have a different pattern compared to the first page of results, which will ruin the CTR comparison model.
- Remove any search queries with less than 30 clicks or so – Without statistically significant click data, the measured CTR can be reported inaccurately. It may appear much higher or lower than it truly is, due to random fluctuations. A good rule of thumb is to remove results with less than 30 clicks but to be more accurate you may wish to remove results under say 100 clicks, depending on how popular the website is.
Now we have filtered out the significant, non-branded, page one search queries, we can plot them on a scatter x-y chart. Select just the “Position” column and plot that against the “CTR” column with positions set on the bottom x-axis:
Add a trend line to this data, which is usually done by right clicking on of the data points. None of the trend lines really fits the predicted CTR model for search engine rankings perfectly, but using an exponential or power based trend line is a close fit:
By ticking the “Display Equation on chart” box you can also see the equation for the trend line which we can use later. Now it’s clear which results are better than average and which ones are below than average:
Using the equation you can work out the expected CTR of each position based on the website’s average:
Now you can simply work out the difference between the recorded CTR and the expected CTR to see if it’s over or under performing:
This data can now be sorted to work on the search queries with the most clicks, most impressions, highest rankings or with the worst CTR difference:
Using the CTR Analysis Data
The most important search queries to start looking at are the queries with a high potential for organic traffic. That’s ones which have a high number of clicks or impressions, with a lower than expected CTR difference.
Once you’ve identified the problematic search queries, check which landing pages are triggered by the queries and ask yourself how relevant the page title and description snippet below are for the search query you’ve just searched for:
Remember that if there is a poor or missing meta description on a landing page, then Google will automatically create one from the rest of the page contents. This automatically generated description may have a better or worse CTR than the text on the description page:
Think of a page title that connects with your audience on an emotional level. Also improve the meta description to answer questions, talk about benefits and mention what makes your website stand out above the rest. Rich snippets also make listings stand out by adding review ratings and additional information taking up more vertical space:
It doesn’t matter how well a website is doing, there’s always room for improvement when it comes to CTRs and you’ll always find listings which aren’t doing as well as they could be. This process can be repeated every few months to constantly improve the lowest performers and raise the overall CTR of a website.
Overall Effects of Improving CTR
By improving your search listings’ CTR you’ll get these key benefits:
- Increased free web traffic – Without paying for ads or performing SEO, you can still raise the number of visitors.
- Improved Google rankings – Google now considers CTR as a ranking factor, so a good CTR can also increase your presence.
- A stronger brand presence – Even if you don’t get more clicks, people will notice your listing more and therefore your brand name.
Don’t stop at your current page titles and meta descriptions, which may have been written years ago, make the small effort to improve your CTR and see how much it impacts your bottom line!
— Jonathan Ellins (@Jonathan_Ellins) 16 November 2016