Online ads have become common practice for political campaigning. In the business world, the possibility of tailoring social media ads to specific target groups such as women, young consumers or specific regions is seen as a major ingredient of efficient persuasion. While prior research has focused on online campaigning more generally, we have very little knowledge about the effectiveness of tailored messages. We argue that political messages tailored to the specific needs and priorities of specific voter groups is a highly promising strategy in political campaigning. To empirically test our argument, we fielded a large-scale field experiment in cooperation with the German Green Party. More specifically, we ran a social media campaign targeting middle-aged women in the context of the European Parliament Election in 2024. Randomizing at the ZIP-code level, the campaign delivered messages that were specifically tailored to middle-aged women in the treatment group, while users in the control group only received generic campaign ads. Overall, the campaign reached 686,000 women. Our results show that neither aggregate-level turnout nor the vote share of the Greens has been affected by the tailored campaign. A post-election survey fielded among middle-aged women additionally indicates that ad reach and recall in the target population were very low. The results show that tailored online campaigning does not have the impact that commentators usually claim, as (1) it is very difficult to reach voters in the target group and (2) even if voters have been reached, tailored online ads do not affect voting behavior.