In part one, The Deception of Averages – In Direct Response Rates, we talked about how campaign managers can inadvertently lose sight of performing variables simply because they fall below the average response rate.
In this next part, we’ll discuss how some non-branded keywords can be labeled as “under-performing” potentially causing a entire ad word campaign to be unnecessarily shut down.
“Non-branded campaigns don’t work! Shut them all off!”
We often hear this type of hyperbole thrown by executives towards PPC Managers and the agencies they work with. The view of performance that often precedes this unpleasant exchange is typically a cursory evaluation of campaign (or even worse) account level performance. They can’t help but notice that their non-branded campaigns are the ones dragging down the overall account average—and thus their wrath on these campaigns is unleashed!
As comforting as this decision might seem, the problem of course is that there are always some high-performing, non-branded keywords in these campaigns and, more importantly, some non-branded keywords that never got the chance to be measured as good or bad.
So how does a keyword get labeled as under-performing when it never got the chance to be seen? This occurs when too many non-branded keywords are grouped together in an ad group fighting for the same budget under the same campaign. At an average level, the ad group looks to be under-performing, which hurts campaign performance. Look closer, however, at the ad groups and their keywords and you will notice certain keywords never received clicks and so of course didn’t produce conversions. This sort of “incomplete” grade occurs when potentially effective keywords are in groups with high-traffic keywords that soak up the ad group’s budget. They never got a chance to be seen!
So how do we give these keywords a fair chance to be seen? We’ve had success in identifying these non-branded keywords, pulling them out and giving them their own ad group, campaign and budget. A variant of the classic “peel & stick” PPC move, normally applied to over-performing keywords now applied to those with that “incomplete” grade.
Executing this re-segmentation should provide opportunity for these keywords to be seen and clicked, thus giving them a chance to generate conversions. More budget to the strong performers and no mercy for the under-performers. One of the best antidotes to the deception of averages in non-branded search is to dig into keyword performance and segment out those over-achievers and optimize their budget. Note that this will optimize the performance of your non-branded PPC campaigns, not necessarily the hasty decision making of your company’s executives!
Senior Marketing Analyst
Look for our final installment of this three part blog series, The Deception of Averages – In Customer Frequency Testing, coming on November 22nd!