Bidding & Optimization Strategies for the Modern Search Marketer
Get the guide today.
In July 2010 Sembox agency executed the research on the ad text efficiency in PPC campaigns.
The research was based on analyzing the performance of ad texts used in PPC campaigns of Sembox clients. The correlation between ad text ranking and CTR was the subject of a detailed analysis, while the other campaign settings remained unchanged.
What the research managed to establish is that:
- There is a significant correlation between CTR and text tanking of the particular ad text – the average CTR at 11th spot is ca.1% and increases to achieve the CTR of 8% at the top place.
- The performance of the ad text at spots 7-9 is very similar (1,5 – 2,2%). It is probably due to the relatively low exposure of the ad among other ad texts being presented at the same time.
- The performance of the ad text at spots 1-6 depends strongly on ad text ranking and there can be seen a significant improvement between the following spots of the ad text (from 0,5% up to 2,5%).
Pic.1. The correlation between average CTR and ad text ranking.
A more detailed research was executed for a sample of the ad texts meeting the qualitative criteria and the results were as follows:
- The ad text shift from 11th1 to 10th spot (bar no. 10 at the chart below) brings the relative drop of CTR by 6,5% - it is probably connected with moving the ad text from the last (utmost) spot to a middle one which is simply less exposed.
- The ad text shift at 8th and 9th spot (ie. from 9th to 8th and from 10th to 9th) brings insignificant changes in CTR – because of reasons provided at the previous page.
- The ad text shift in spot range of 3-7 (ie. from 4th to 3rd, from 5th to 4th, etc.) brings a substantial difference in CTR – on average by 33-39% which means that it may be reasonable to obtain the highest possible ad text ranking in this range that will result in a larger volume of clicks (visits) at the campaign landing page.
- The ad text shift from 3rd to 2nd spot and from 2nd to 1st spot (bars no. 2 & 1 at the chart below) results in insignificant change in CTR – 17,3% and 8,2% accordingly which proves a comparable exposure of ad texts at top spots.
Pic.2. The relative % change of CTR between adjoining spots of the same ad text.
Reversing the conclusions presented at the previous page we can say that:
- A sample ad text that has a CTR of 1% at 11th spot will be 7 times more effective when distributed at the top spot – the average growth of CTR for all analyzed ad texts was equal to 707,6%.
- The increase in ad text performance can be more significant and depends on base efficiency of ads – the sample of ad texts that were significantly more efficient than 1% at 11th spot proved to achieve even 9 times higher CTR at the top spot.
Pic.3. A simulation of CTR of the ad text that has a base CTR of 1% at 11th spot.
The relation between ad text ranking and CTR
Almost everyone who is involved in PPC campaign management is aware that ad text ranking influences campaign effectiveness. The click-through ratio (CTR) is higher if the ad text is more exposed (ie. closer to the top of the page) to user’s eyes. We tried to calculate the strength of this relation in our research. In order to do this we analyzed the average daily rankings of ad texts used in campaigns over several months (the detailed methodology of the research is described in the last part of this document).
The following groups of ad text were analyzed:
- all texts that had any number of views/clicks (marked as “0/0″ segment),
- ad texts that had at least 200 views and 10 clicks (“200/10″ segment),
- ad texts that had at least 1000 views and 20 clicks (“1000/20″ segment),
- ad texts that had at least 1000 views and 100 clicks (“1000/100″ segment).
We would expect to observe significant differences in effectiveness of ad texts belonging to the most general segment (“0 / 0″) because of the fact that a lot of texts had statistically insignificant results (to few impressions or clicks) and the best performing ad texts from “1000 / 100″ segment which should have been the most valuable from the statistical point of view.
As a result, we observed the following performance:
|Segment / Rank||1||2||3||4||5||6||7||8||9||10||11|
|0 / 0||7,73||5,54||4,51||3,49||2,81||2,30||1,69||1,49||1,69||1,19||0,45|
|200 / 10||7,83||5,58||4,53||3,53||2,85||2,34||1,75||1,62||1,97||1,47||0,69|
|1000 / 20||7,93||5,59||4,54||3,54||2,86||2,36||1,76||1,63||2,00||1,53||0,77|
|1000 / 100||8,14||5,66||4,62||3,64||2,95||2,42||1,88||1,92||2,16||1,75||1,12|
Tab.1. Relation between CTR and ad text ranking in various segments.
The difference in performance of ad texts belonging to specific segments is insignificant at the same positions (e.g. 7,73% – 8,14% for the 1st spot). What is more important, the CTR change trend across all rankings is very similar in all segments. This is presented at the chart below which visualizes figures shown in Table 1:
Pic.4. Relation between CTR and ad text ranking in various segments.
The detailed analysis of particular segments enabled us to identify exceptions which made the whole summary slightly inaccurate – some segments contained ad texts which had achieved different performance results (varying from 2% to 11%). We believe this was particularly a result of:
- specifications of campaign subject,
- competitiveness– number of ads exposed to the user at the same time,
- quality of ad texts of competitors and relative attractiveness of ads being the subject of this research.
These are the main reasons why the average (absolute) value may not accurately express the differences of effectiveness of ad texts. That is also why we decided to conduct a more detailed analysis which is also described in the later part of this research.
Should you want to find out the results of some further research done by the Sembox agency, please see the link below: