Google recently announced that it would further limit its election ads to

• audience targeting based on age, gender, and general location (postal code level)
• context targeting (i.e. showing ads based on the content being viewed)

Up to this point, the application of predictive modeling to “microtarget” individuals or small groups of individuals, well-entrenched in the digital marketing sphere, had been advancing in the political realm as well.  Predictive models have been used to better target political ads for at least 15 years. This targeting has become increasingly sophisticated and micro, and Brad Parscale, who handled digital marketing for the Trump campaign in 2016, described how his algorithms would create tens of thousands of different Trump ads daily on Facebook, to determine what specific presentations should be shown to which voters.  Parscale gave this Facebook campaign the credit for the Trump victory.  

Google says that it never allowed “granular microtargeting” of political ads, and that its recent step is a simply a further tightening.