New changes will force marketers to embrace machine learning, giving Google ultimate control over campaigns.

 
   Adept Insights
  • Google recently made changes to its Search Terms Report to only include terms that a significant number of users searched for.
  • We believe the change will eventually cause more marketers to use Google's machine learning automation, giving Google ultimate control over campaigns.
  • Before 2019, Google's machine learning automation rarely outperform our manual automation. However, the tool has become more intelligent at prediction. 
  • We hope machine learning will offer some fresh insights that yield short-term positive results but we fear this change will have negative long-term impacts.

 

Google made the decision this fall to hide significantly more paid search query data than ever before. In a statement provided by the company, they claim it is to improve user privacy.

"In order to maintain our standards of privacy and strengthen our protections around user data, we have made changes to our Search Terms Report to only include terms that a significant number of users searched for. We're continuing to invest in new and efficient ways to share insights that enable advertisers to make critical business decisions."

Behind the Changes

While user privacy is a legitimate concern, we believe declining revenues may be a contributor as well. Google must protect themselves as statewide user privacy laws take effect. However, Google is also seeing shrinking ad revenue growth, primarily due to Amazon and COVID-19. In 2020, Google's ad revenues dropped for the first time ever, and it is projected Google will post a 5.3 percent (down to $39.58 billion) in 2020. That brings Google's share of the U.S. digital ad market to 29.4 percent, down from 31.6 percent last year. As a result, Google is accelerating the adoption of machine learning, which kills two birds with one stone. User privacy improves and Google gains more control of the market.

By hiding search query data, advertisers lose the ability to optimize campaign budgets, producing more Google revenues. Specifically, there will now be about $27 thousand worth of search term data left unseen for every $100 thousand spent. That makes optimization (specifically adding negative keywords or new long-tail positive keywords) more complicated and, in some cases, not possible. Also, marketers may receive irrelevant clicks in search and not know it, negatively impacting paid search performance.

The Future of Machine Learning

The change also forces marketers to use Google's machine learning automation (see Google Analytics 4). For more than two years, Google's made updates designed to increase control through machine learning, making it a 'best practice' resulting in increased adoption rates. But by making the platform hands-off to marketers, Google now gets to make all of the decisions, giving them ultimate control over campaigns and more money in their pocket. 

At Adept, we have tested the effectiveness of machine learning bidding methodologies on all campaigns. Before 2019, Google's machine learning automation rarely outperformed our manual automation. However, the tool has become more intelligent at prediction, allowing us to run 85 percent of campaigns in 2020 on some type of automation because those automations outperformed our manual optimization. The improvement gives us lukewarm confidence that the eventual switch to machine learning will have minimal impacts on campaigns, at least in the immediate future. 

Embracing Machine Learning—For Now

While it would be nice not to be forced to make changes, marketers have little choice but to accept Google's search query updates. Perhaps machine learning will offer some fresh insights over time that will yield positive results. Or, we'll have to wait until another search engine, say from Apple, disrupts the scene and forces Google to rethink their efforts.

If you have questions about Google's recent changes and your campaigns, contact us.

[Read More]