Can you still split test on Adwords?
After not using Adwords for a couple of years; all I wanted to do was create a campaign to test domains names for a new website.
After unnecessary blood, sweat, tears the Adwords campaign was started. One campaign was made with 1 adgroup containing 9 different ads. Each ads was identical except for the display url which was the domain names that were tested.
Here is the problem. The campaign was set to rotate each ad evenly, indefinitely.
However, in spite of that, the ads were not rotated evenly. In fact, the ads that got the most clicks received almost 3 times more impressions than the ads that got the least amount of clicks.
This invalidates the test since I don't know whether the ads with the most clicks were because they were actually the best or got more impression.
To counteract this, a new campaign was started were each ad was placed in a separate ad group. I will check the results later tonight, however this forces more non-needed complications.
They are some stats below from the original campaign:
EDIT to report on new campaign.
The new campaign mentioned before was worst. Although each ad was placed in a separate adgroup, only the ad that received 20% impressions and 8 clicks from above received any traffic. All other adgroups got 0 impressions.
After unnecessary blood, sweat, tears the Adwords campaign was started. One campaign was made with 1 adgroup containing 9 different ads. Each ads was identical except for the display url which was the domain names that were tested.
Here is the problem. The campaign was set to rotate each ad evenly, indefinitely.
However, in spite of that, the ads were not rotated evenly. In fact, the ads that got the most clicks received almost 3 times more impressions than the ads that got the least amount of clicks.
This invalidates the test since I don't know whether the ads with the most clicks were because they were actually the best or got more impression.
To counteract this, a new campaign was started were each ad was placed in a separate ad group. I will check the results later tonight, however this forces more non-needed complications.
They are some stats below from the original campaign:
Code:
percent clicks impressions 20.00% 8 154 19.61% 5 151 17.27% 6 133 7.92% 0 61 7.40% 0 57 7.14% 0 55 6.88% 1 53 6.88% 0 53 6.88% 0 53 SUM 20 770
The new campaign mentioned before was worst. Although each ad was placed in a separate adgroup, only the ad that received 20% impressions and 8 clicks from above received any traffic. All other adgroups got 0 impressions.