Google Introduces Machine Learning for Age Estimation in Advertising
Google recently informed advertisers and publishers about an exciting shift in their ad personalization strategy. Starting soon, they’ll use machine learning to estimate the ages of signed-in users in the United States. This move is all about safeguarding young people while they engage with Google products.
What’s Changing?
Here’s the scoop:
- Ad Personalization Disabled for Minors: Google will disable ad personalization for users under 18, ensuring that sensitive content is not shown to younger audiences.
- Sensitive Creative Categories: Advertisements that fall under sensitive categories will also be restricted for these users.
This change was previously announced, but the rollout is imminent, targeting a “small set of users” in the U.S. over the next few weeks, allowing Google to monitor the impact before expanding further.
Insights from Google
In an email to advertisers, Google elaborated on how this will work:
“As previously announced, Google will begin using machine learning to estimate the age of signed-in users in the United States. Over the next few weeks, we’ll begin to roll out this update to a small set of users in the U.S. to help us further protect young people as they use Google products. We’ll closely monitor this before we roll it out more widely.”
Key Protective Measures
- Ad Safeguards for users flagged by the machine learning model as likely under 18, several ad protections will kick in, including:
- Disabling ad personalization
- Disallowing sensitive creative categories from serving
These updates will be implemented across Google’s publisher ad products such as Ad Manager, AdSense, and AdMob when account information is utilized. Advertisers don’t need to take any immediate action in response to these changes.
Technical Details: How Does It Work?
Now, how does Google figure out a user’s age? They utilize a blend of:
- Age Estimation: This involves analyzing various signals tied to a user’s account, like search history and YouTube viewing habits. These insights help decide if a user is likely under or over 18.
- Age Verification: If users believe the machine learning model has incorrectly categorized them, they can correct their age. Options include uploading a photo of a government ID or a selfie to verify their age.
Industry Reaction
This announcement has sparked discussions among advertisers. For instance, Drew Cannon shared an early reaction on social media, expressing concern about potential policy violations, adding a bit of humor to the situation with:
“Here come the policy violations flags. @GoogleAds and publishers. This should get interesting #ppcchat”
As marketers, these changes bring both challenges and new considerations. It’s vital to stay informed and adapt strategies accordingly as Google navigates this new frontier in user privacy and age estimation.
Stay tuned for more updates as Google rolls this out!
