Targeting on Google Display Network has been a hot topic the last two weeks. I keep hearing confusion about performance of managed placements once they’re moved from contextual campaigns to placement targeted campaigns, particularly in that they receive lower traffic volume and have a higher cost per click.
In order to understand what’s happening here, we’ve gotta go back in time a few UI iterations to the days when we saw two additional targeting settings within a campaign opted in to the GDN. Do you guys remember “Show ads on pages that match the broadest targeting method” (broad reach) and “Show ads only on pages that match all selected targeting methods” (specific reach)?
Back when we had these settings we would need to set them according to the type of targeting we were after. If we were interested in a contextual target, we would select the first option. If we were interested in only targeting certain placements, then we would select the second option. We would also select the second option if we were interested in showing our ads alongside content that matched a contextual theme but only on certain placements. You old timers like me may recall this used to be called an enhanced campaign. Way to recycle terminology Google.
The current interface does not have either of these options. We’re simply allowed to opt in to the GDN for all features or select something more specific like Remarketing or Mobile apps. Now that the old broad and specific reach options have been removed, Google informs us via the help center that:
Contextual targeting is always used when an ad group has keywords and the Display Network is selected on the Settings tab.
This means that any ad group containing both placements and keywords will work according to the following example, also from the AdWords help center:
Alrighty, so now that we’ve reviewed settings, these are important when optimizing your contextually targeted campaigns on the GDN. Many account managers will start out by casting a wide net with a contextually targeted campaign. Which is then sculpted and sliced into a placement targeted campaign or several campaigns with a mix of targets.
The problem we run into is that if placements found via a contextually targeted campaign (automatic placements) are simply converted to managed placements in that same ad group. We actually fall into the cupcake scenario described above. Keywords are still present and our targeting now changes from targeting content across the GDN that matches our keywords to targeting only that specific placement and only when the content on that placement matches our keywords.
This dual targeting result s in a significant restriction in available ad space for our ads and volumes will decline. Because there is increasingly limited ad space available on that page, we usually only see a couple ad spots on a single page, but we’ve also told Google we’re only interested in those placements when the content on that pages matches our contextual theme which places us into an interesting auction in which we will need to beat out all other advertisers that may qualify for this ad placement and we often see increased CPC. For more information on the increase in CPC, I recommend checking out an older article from John Lee on Placement Targeted Ads.
I recommend moving these managed placements either into separate ad groups or campaigns. I usually move them to a new campaign so I can clearly see where my placements are located. From there you can cross reference site exclusions for controlling ad serving like you would with negative keywords and match types in a search campaign.
Hopefully this helps clarify some confusion