The Power of AI: Driving Results and Saving Resources in a Competitive School Board Election

Ad campaigns were created 8X faster with BattlegroundAI and lowered costs by 18% compared to using human power content alone.

‍

Down-ballot races like School Board face substantial time and resource constraints. It is critical that these campaigns experiment with innovative strategies and tactics as they can have an outsized impact on election results. 

‍

The opportunity occurred in a crowded race. The team only had a few weeks left to increase name recognition and make an impact on voters. They needed to develop and deploy an expansive, high-impact paid digital advertising campaign. Below outlines the four core steps to building a successful paid media campaign, quickly and delivering results.

‍

Step 1: Polling and Research

The first step was to understand the existing research and polling results. A poll by Change Research conducted in the district showed top messages and audiences. Notably, 77% of voters had not heard of the candidate before, indicating a significant amount of work ahead.

‍

Step 2: Media Planning

The budget was divided among platforms based on our audience demographics and targets, which included Meta, Programmatic, Google Display, and Google Search ads. Although we did not have video content developed in time, we utilized existing imagery and photo assets.

‍

Step 3: Content Development

• Developing a typical  Meta, Google Search and Display campaign (all with different copy requirements) requires about 800 words. Before the help of AI this content was developed  based on a combination of intuition and research and it took about 40 minutes. Typically a campaign may have many sets of copy for each audience or policy issue, and regularly updated or refreshed

• With the help of AI: BattlegroundAI was used to quickly and effectively create content fine tuned to the audience, message and platform. The process involved the following steps

‍

1. Create an Audience: Define the primary target audience based on demographics.
2. Build a Campaign: Select each platform, objective, and policy topic or messaging point.
3. Upload Files: Provide a CSV file of our research to the platform.
4. Generate Content: Create hundreds of message variations customized to fit each advertising platform's parameters in minutes using AI.
5. Rinse and Repeat: Steps 1 - 4  for each audience, platform, and message.
6. Review & Approve: Review the results with the communications team and launch the campaign.

‍

Step 4: Results

The campaign resulted in over 11 million impressions. Advertising content generated using AI (and reviewed by humans) resulted in an 18% decrease in Facebook CPMs compared to human generated content, allowing us to optimize our budget and even expand to digital out-of-home advertising in the days before Election Day. It also saved hours of time, with the help of the platform it took about 5 minutes or less to write an ad set, which is about 8X faster.

‍

‍

This allowed for more resources to go into strategy and optimization. Although we didn’t win, we still gained valuable insights and demonstrated the ability to scale a paid digital campaign faster and more efficiently than ever before.