A couple of weeks ago we had a look at possible traffic losses when duplicating a campaign. It turned out that the Quality Score can be used as an indicator for when traffic can be shifted from the old campaign to the new one at a minimum loss of Impressions and Clicks. Please view our previous post for more information.
We noticed in the past that it makes a difference if you upload new campaigns in a mixed account that contains both Generic and Brand keywords or in a pure Generic account. To estimate the impact we decided to run a second test series.
For this second set of tests we chose high traffic Generic campaigns from pure Generic accounts that were duplicated twice into two separate, completely new accounts. The settings were identical, the only difference was that the shop name was added as phrase negative in only one of the new accounts.
To avoid cannibalisation the campaigns, including the original one, ran on an alternating two hour schedule.
Traffic Differs Significantly
Comparing the performance of the test campaigns of the first set of campaigns we can see a significant difference in traffic.
Even though Quality Scores are nearly identical in all three campaigns, the campaign without the Brand Negative received by far the highest amount of traffic, even more than the original campaign.
Interestingly, out of more than 17K Impressions there was only one Impression from a search term containing the brand name. So the difference is not directly caused by the missing brand impressions. To further investigate, we have set up a second test with the same conditions.
The second set of campaigns shows similar results:
Again, the negative brand name is the only difference between the two new duplicate campaigns. Both ran intermittently.
In contrast to the first set, here the campaigns in the accounts without negatives do not have a significantly higher number of Impressions than the original ones but they still received much more traffic than the campaigns in the account with the shop name as Negative while the Quality Scores are more or less identical.
Here too, the number of impressions generated by brand related queries was negligibly low. The gap in impressions seems to be an indirect result of the brand negative.
Conclusion: Wait Before Implementing Brand Negatives For New Campaigns
Both results suggest the conclusion that adding the shop name as Negative to some campaigns, while the URLs contain the same term, has a negative influence on the ad rank, leading to fewer Impressions and therefore a loss of traffic. You could also argue that the few brand impressions with strong CTR help to build account relevancy.
To avoid a loss in traffic we would suggest to wait for a few days before adding your Brand term as Negative to your new Generic account. The amount of actual Brand searches accidentally generated will be very low but you can avoid losing a significant amount of traffic.