Google is turning to artificial intelligence (AI) to boost its advertising game and make up for lost ad revenue, following Apple's privacy changes that have made it harder to target ads to users online.
At the heart of this strategy is Performance Max, an AI-driven tool launched two years ago. It's smart at figuring out the best places to show ads across Google's own platforms, like YouTube, Google Search, and various websites where Google displays ads. Lately, Google's been encouraging even small and mid-sized businesses to give Performance Max a try.
Retailers, in particular, are finding value in Performance Max. It helps them find and attract customers, leading to increased sales—a fact backed by ad agencies in the know. For example, Tinuiti, an ad agency, observed that a whopping 90% of Google shopping ads for retailers effectively used Performance Max in the third quarter. This has led to a 16% jump in retailer ad spending year-over-year with Google.
And it's not just Tinuiti that's investing more in Performance Max. Belardi Wong reported that 68% of its Google ad buys for retailers are through this tool, up from 50% last year. Digital agency Arm Candy also noted that almost half of their Google ad budget for retailers goes into Performance Max. Another agency mentioned a 60% spend on this tool for Google shopping ads.
However, not everyone is convinced. As Performance Max becomes more common, some advertisers are questioning how much they can trust AI to deliver better ad performance without showing them how it's done or where the ads end up.
For instance, Arm Candy plans to reduce its spending on Performance Max this year. Their main concern is the lack of control over where ads are placed, making it tricky to track results. Performance Max doesn't always buy the same type of ads consistently—it might choose more YouTube ads sometimes and more display ads another time.
In response, Arm Candy is shifting funds back to traditional, human-managed campaigns. They've even created an internal tool to analyze ad placements more closely.
Others have noticed dips in performance too. For example, Merkle, a digital agency, reported lower results after using Performance Max. Yet, Brendon Kraham from Google explained that Performance Max aims for a diverse mix of ad types and that any drops in performance could be due to the availability of ad spaces.
Amidst these concerns, Google has been working on tweaks to give advertisers more say in how Performance Max works, like excluding certain keywords or focusing on acquiring new customers.
Despite the skepticism, some agencies, like Belardi Wong, continue to use Performance Max. They find it valuable in reaching potential customers early in the shopping process, even if it means a little less control over the campaigns.
In a nutshell, Google's Performance Max is using AI to adapt to a privacy-focused online world, aiming to deliver effective ad campaigns. While it offers promising results for many, it also poses challenges for advertisers who prefer hands-on control and clear insights into their ad placements.
Read next: New Chinese AI Outperforms Meta LLM With Half the Parameters
At the heart of this strategy is Performance Max, an AI-driven tool launched two years ago. It's smart at figuring out the best places to show ads across Google's own platforms, like YouTube, Google Search, and various websites where Google displays ads. Lately, Google's been encouraging even small and mid-sized businesses to give Performance Max a try.
Retailers, in particular, are finding value in Performance Max. It helps them find and attract customers, leading to increased sales—a fact backed by ad agencies in the know. For example, Tinuiti, an ad agency, observed that a whopping 90% of Google shopping ads for retailers effectively used Performance Max in the third quarter. This has led to a 16% jump in retailer ad spending year-over-year with Google.
And it's not just Tinuiti that's investing more in Performance Max. Belardi Wong reported that 68% of its Google ad buys for retailers are through this tool, up from 50% last year. Digital agency Arm Candy also noted that almost half of their Google ad budget for retailers goes into Performance Max. Another agency mentioned a 60% spend on this tool for Google shopping ads.
However, not everyone is convinced. As Performance Max becomes more common, some advertisers are questioning how much they can trust AI to deliver better ad performance without showing them how it's done or where the ads end up.
For instance, Arm Candy plans to reduce its spending on Performance Max this year. Their main concern is the lack of control over where ads are placed, making it tricky to track results. Performance Max doesn't always buy the same type of ads consistently—it might choose more YouTube ads sometimes and more display ads another time.
In response, Arm Candy is shifting funds back to traditional, human-managed campaigns. They've even created an internal tool to analyze ad placements more closely.
Others have noticed dips in performance too. For example, Merkle, a digital agency, reported lower results after using Performance Max. Yet, Brendon Kraham from Google explained that Performance Max aims for a diverse mix of ad types and that any drops in performance could be due to the availability of ad spaces.
Amidst these concerns, Google has been working on tweaks to give advertisers more say in how Performance Max works, like excluding certain keywords or focusing on acquiring new customers.
Despite the skepticism, some agencies, like Belardi Wong, continue to use Performance Max. They find it valuable in reaching potential customers early in the shopping process, even if it means a little less control over the campaigns.
In a nutshell, Google's Performance Max is using AI to adapt to a privacy-focused online world, aiming to deliver effective ad campaigns. While it offers promising results for many, it also poses challenges for advertisers who prefer hands-on control and clear insights into their ad placements.
Read next: New Chinese AI Outperforms Meta LLM With Half the Parameters
