About /
Clients are fact-seekers, and we can’t blame them. Every day we face a tug-of-war between ‘what clients want to know’ and ‘what we can actually tell them in confidence.’ It is our responsibility as strategists to present clients with data that brings them value. However, with a constantly changing search engine landscape, it has become increasingly difficult to convey ‘how much’ or ‘how many.’ This POV is aimed at establishing a general click-through-rate (CTR) to help clients understand the impact of organic rankings within SERPs.
According to searchenginewatch.com, CTR is defined as the rate at which users click on an ad, expressed as a percentage. This is calculated by dividing the total number of clicks by the total number of impressions (or the total number of times the listing is displayed).
Many attempts have been made to establish a CTR scale that accurately depicts consumer behavior, but different attempts have yielded different results. Rosetta focused on four studies that spanned a variety of seasons, markets, buyer types, and study durations. They are as follows:
In addition to these studies, our team considered incorporating a few others. For example, Cornell University conducted a CTR study in 2004, but our team concluded that this “controlled” study did not adequately predict true user intentions. As a result, we ultimately decided against incorporating additional studies like this because we believe it would have compromised the comparative nature of our data.
| Rank | CTR |
|---|---|
| 1 | 29.01% |
| 2 | 12.82% |
| 3 | 9.21% |
| 4 | 6.38% |
| 5 | 4.85% |
| 6 | 4.01% |
| 7 | 3.46% |
| 8 | 2.99% |
| 9 | 2.54% |
| 10 | 2.38% |
| 11 | 2.18% |
| 12 | 1.53% |
| 13 | 1.27% |
| 14 | 0.93% |
| 15 | 1.00% |
| 16 | 1.18% |
| 17 | 1.30% |
| 18 | 1.15% |
| 19 | 1.23% |
| 20 | 1.49% |
In order to account for differences in seasonality, study duration, and methodology, our team took a simple average of each study’s 1-10 CTRs to develop our own CTR chart for positions 1-10.
When developing CTRs for positions 11-20, the data were sparse. Because the Optify study was the only study to report CTR data on positions 11-20 on Google, Rosetta decided to use these data as a base to predict what the remaining studies would have looked like if they had collected through the top 20 as well.
To do this, we calculated the relative differences between the Optify study and the remaining three studies through the first 10 positions. We then applied the degree of difference to the corresponding rank on the second page in order to ‘simulate’ the remaining three studies. For example, Slingshot SEO’s CTR for a number one rank was exactly half that of Optify’s (Figure 1).
| Rank | Optify (Dec 2010) | Slingshot (Jan-July 2011) |
|---|---|---|
| 1 | 36.4% | 18.2% |
To develop a predictive rank for the Slingshot study, we applied this same degree of difference to the corresponding second page placement (Figure 2).
| Rank | Optify (Dec 2010) | Slingshot (Prediction) |
|---|---|---|
| 11 | 2.6% | 1.3% |
We then used this methodology to simulate 11-20 CTRs for the Slingshot study, Enquiro study, and Chitika study.
Lastly, we placed greater weight on Optify’s 11-20 data to account for the fact that the simulated studies were merely predictions and developed average values for 11-20 CTRs.
Every search yields different results, so some important factors play into the dynamic nature of CTRs.
As part of Rosetta’s effort to stay abreast with current best practices, identifying factors that influence CTRs is of utmost importance to our team. We are currently testing this at Rosetta.
One way Rosetta is testing this is by looking into the effects of altering the presence of snippets. By gearing ‘calls-to-action’ more toward the target market(s), the average CTRs that we presented will likely be affected. We estimate that better snippets will yield higher CTR regardless of position within the top 5. We can also test the impact of our work against CTRs by shifting the brand placement, “sales-focused” copy, and the introduction of attention-grabbing words like “free.” Using Google Webmaster tools,* we can measure the different factors that impact click distribution.
* Google Webmaster Tools gives only an estimate and will always be experimental in nature.