John Whipple - Guest Blogger


User Advice for Sweeping Underperforming Ads from Your PPC Campaigns

John Whipple (@john_whipple) is a Strategic Account Manager at BoostCTR, the world’s market place for performance marketing creative.

In User Advice on Rotation Settings Best Practices, I examined some of the benefits of moving your campaigns from optimized to even rotation, especially if you are considering optimizing your PPC account through ad copy testing. However, before making this settings change it is important to check how many ads you have running in the ad groups that will be affected by this change. As is often the case when accounts are running optimized rotation, the number of ads in your ad groups can grow quickly if you do not pay attention to it. This brings us to the second topic of this series: Cleaning out underperforming ads (what I refer to as Ad Sweeping).

A wise objection one might have to moving their campaigns from uneven to even rotation is that it could lead to a serious dip in performance. And this concern is valid if you have not paid attention to how many ads are running in your ad groups while uneven rotation is in place. Given that even rotation is the ideal environment for ad testing, it is important to have a methodology in place for removing bad ads from your ad groups prior to making this change.

Pruning your ad groups of its worst ads is one of the lowest hanging fruit for optimizing your account. While there is no magic number of ads you should be running in your ad groups, there are some general best practices you can use to help understand which ads are hurting your account. Having a clear goal and set of metrics defined is the first step. Knowing your key success metrics will help you make the best decisions on which ads to pause and remove subjectivity from the process. For our purposes, I have included an example below with the simple goal of improving the ad group CTR:

Ad Groups_Search Advertising Spring Cleaning

With your success metrics defined, you can start to identify and eliminate the most obvious under performers. In Table 1, the ad group is currently running 5 ads with one clear top performer. In Table 2, the analysis demonstrates how you can compare ads using underperformance on a success metric and statistical confidence to identify which ads make the most sense to pause. If you do not have the tools to run statistical analysis, then you should establish threshold criteria such as looking at ads over a comparative time frame of at least 10 days, where each ad has 1000+ impressions.

Using this kind of methodology will help you objectively decide which ads to keep running and which to remove. By pausing your worst underperforming ads, not only do you improve the average level performance of the ad group, but you also use the impression volume in your ad groups more efficiently by getting more searchers looking at your best ads—and all of this without having to write a single ad!

Key takeaway - it is important to remove, or sweep, lower performing ads that could be stealing impressions from higher performers.

Now that we have established the best practices for rotation settings, and have a strategy in place for reducing the number of ads running in your ad groups, the stage is set to run effective ad copy tests that can yield strong results. Look out for the final article in this series where I will examine some best practices for ad copy testing.

Keep reading

User Advice on Rotation Settings Best Practices - Part I

Maintaining effective, ROI-positive paid search campaigns can be a tedious task, mired in complicated metrics and methodologies. At BoostCTR, our philosophy is that optimization of a PPC account revolves around three pillars:

  • Bid management
  • Keyword relevance
  • Ad management

In this series, I will focus on various aspects of ad management and how the time you devote to this task can yield tremendously positive results. The series will break down into three sections: 1) Rotation settings best practices, 2) Cleaning out underperforming ads (Ad Sweeping), 3) Ad copy testing, and for this article we will focus on the first.

1. Rotation Settings

Ad Rotation When I speak with online marketers and SEM experts about ad management, the first thing that comes to mind is ad copy testing– and rightfully so. However, before you think about testing, you should consider what rotation settings are in place throughout your account. While most platforms default to click-optimized (uneven) rotation, this is not an ideal environment for testing. Giving the lion’s share of impressions to only one ad lowers confidence in the test results and requires your tests to run longer before a conclusion can be made. Statistical significance is a useful measure for objectively evaluating the difference in performance between two ads. Even though search engine platforms employ statistical analysis in the algorithm that decides which ad to shift impressions towards, when it comes to testing ad copy yourself, data is your most valuable asset.

Having your campaigns set to even rotation (rotate more evenly in AdWords and Bing Ads) will help each ad receive a sufficient threshold of data to use when making a decision. In an ad group receiving 1000 impression per month, giving 90% or more of the volume to one ad will make the likelihood of achieving statistically significant performance data on the other ad quite difficult; with only 100 total impression the data would be too thin to be confident in that ad’s performance. Moreover, a skewed distribution of impressions between ads can have other ancillary effects on ad performance.

One example of this side-effect is how uneven rotation can impact the keyword-ad distribution. New ads that are inserted into an ad group set with uneven rotation can fall victim to being shown on certain keywords more frequently than others. This alone can muddy the waters when trying to determine an ad’s true performance. In the case that an ad is shown on a disproportionate number of lower performing keywords, the ad’s performance will appear lower than it would otherwise if the keyword distribution were more even. The reverse is true if the ad is shown with more high performing keywords. Either way, your understanding of whether or not an ad is winner can be inaccurate and lead to misguided conclusions in the test.

Lastly, one important thing to note when making a change to the rotation settings is that different platforms make this setting at different account levels. In AdWords, the rotations setting are made at the campaign level, meaning any change you make will affect all ad groups in that campaign. In Bing Ads however, the rotations settings reside at the ad group level, which is helpful for testing ad copy in a specific set of ads groups.

Key takeaway - when it comes to ad copy testing, having the ad data distributed more evenly is always preferable.

With your rotation settings properly in place, you are well on your way to running a successful ad copy test! Look out for the next article of this series where I will review some best practices for how to better manage which ads you have running in your account.

Thank you for reading,



For more information about ad rotation, check out Bing Ads Product Manager Ping Jen’s latest articles:

From Bing Ads with Love: Rotate Your Tires & Ads Regularly

From Bing Ads with Love – Ad Rotation Function Has Arrived

From Bing Ads With Love- Introducing the Bing Ads Ad Rotation Function

Keep reading