To Pin Headlines, Or Not To Pin, That Is The Question
Google Ads introduced responsive search ads (RSAs) in early 2019 as a way to serve ads more efficiently through the process of automation. By including up to 15 headlines and 4 descriptions in each ad, advertisers can leverage machine-learning capabilities and automation to more easily A/B test and show more relevant ad copy to users.
For the advertisers who like to maintain more control over their ad copy, advertisers also have the option to pin headlines and descriptions to a specific position within their responsive search ads, which allows for more control and tailored messaging. However, as it’s indicated in the Google Ads platform, pinning assets will usually have a negative impact on the “ad strength” score, which is a measurement of ad quality and relevance that could potentially affect your quality scores, cost-per-click, ranking, and other factors. Therefore, Google Ads wants you to lean into the automation features and allow their machine-learning algorithms to rotate and optimize your RSA assets to show the better-performing combinations.
Now, although it is recommended by Google Ads to leave RSA assets unpinned, the question still remains: Should advertisers maintain more control over their messaging by pinning specific assets or should we learn to trust the power of automation and machine-learning to help optimize our ad copy?
So When Should I Pin My Headlines?
Context, context, context. The campaign strategy is the key factor to consider here before deciding whether or not to use pinned ad assets. Although pinning assets is not recommended by Google Ads itself, there are still a number of instances where this could be beneficial. Branding, promotional, seasonal, or event-based campaigns may require messaging to remain consistent and accurately display CTAs, slogans, offers, and discounts. These are just a few examples of the instances where pinning assets could make sense.
Otherwise, in situations where any number of headlines could work well together, it may be best to let the machine-learning algorithms optimize to show the different ad variations to find the optimal combination – a practice that is known to lead to an increased CTR%.
Our Case Study:
In an effort to gain some deeper insights into this process, the team at Brawn Media conducted a comparative study to evaluate the differences in the CTR% metric for these two approaches. To conduct this test, we implemented one unpinned RSA and one pinned RSA into each ad group that we are running for two separate accounts. After 30 days of A/B testing the pinned vs. unpinned responsive search ads, the results were as follows:
- In the first account, unpinned responsive search ads generated a significantly higher CTR% (click-through rate) compared to pinned RSAs. The unpinned RSAs generated a 13.87% CTR, while the pinned RSAs generated only 8.94% in the same timeframe.
- In the second account, unpinned RSAs generated an 8.93% CTR, while the pinned RSAs generated a 6.61% CTR during the 30-day period. Although the difference was not as significant as in the first example, it still demonstrated meaningful improvements in ad relevance and CTR% when using unpinned RSAs.
Be Methodical & Continue Testing
From our case study, we saw a higher CTR% from the unpinned responsive search ads vs. the pinned responsive search ads, which implies that there may be opportunity for advertisers to leverage the machine-learning algorithms to generate a higher CTR% from their search ads.
Whether or not a campaign requires the implementation of pinned headlines and descriptions entirely depends on the campaign’s strategy and goals, so unfortunately, there may not be a “one answer fits all” type of result from this case study. However, it does still provide a level of statistical analysis and insight that can be used to impact some of our decision making when creating new ads.
Overall, the biggest takeaway from conducting this case study is the power of testing. Even if you aren’t ready to completely shift towards unpinned responsive search ads, it could still be worth testing and monitoring as we move into a world that is full of more automation.