About the Episode
How can we build audiences in our secondary social channels in a meaningful way? That’s what Andrew Cassel, Senior Social Strategist at Middlebury College sought out to answer. He engaged his community in submitting prompts which he then used to generate beautiful AI images. He posted the images on Pinterest, LinkedIn, and other secondary platforms. His results? On the form, 53 prompts were submitted, surpassing his 30 prompt goal. On Pinterest, a 50% audience increase. On LinkedIn, incredibly meaningful engagement with an alumna. While he’s holding on rolling this out more broadly, it’s a great example of how we can do social listening, co-create content with community, and meaningfully grow audiences with one simple tactic.
Key Takeaways
- AI and Community Co-Creation: Crowdsource prompts from your audience and use generative AI to create images that represent the spirit of your institution. It’s not just about the visuals—it’s about building community and telling authentic stories.
- Diversify Platforms Beyond Meta: While Instagram and Facebook dominate in higher ed, platforms like Pinterest, LinkedIn, Reddit, and even Discord can deliver meaningful engagement and reach new audiences.
- Measure and Track Engagement: Use tools like Google Forms, bit.ly links, and platform-specific analytics to track the success of campaigns and adapt them as needed.
- Short Prompts, Big Impact: When using AI tools like MidJourney, concise and descriptive prompts generate the best results. Developing “prompt engineering” skills is essential for maximizing AI tools.
- AI as a Storytelling Tool: Generative AI can create visuals for abstract concepts like memories, feelings, or moments—bringing alumni stories, student experiences, and campus vibes to life in new ways.
The Spark: Combining Generative AI with Audience Engagement
Andrew Cassel’s project was born out of two goals: to experiment with generative AI tools like MidJourney and to grow Middlebury’s presence on lesser-used social platforms. Inspired by a desire to move away from Facebook and Instagram, Andrew created a campaign that centered on his audience’s input.
Using a Google Form, Andrew asked Middlebury’s social followers to describe the college as they would to a friend. These descriptions were then used as prompts for AI tools like MidJourney and DALL-E to generate evocative, artistic images. The campaign showcased how AI could turn user-generated content into engaging visuals while fostering community participation.
Platform Experimentation: Beyond the Usual Suspects
Instead of relying on Facebook and Instagram, Andrew focused on platforms like Pinterest, LinkedIn, Reddit, and even Tumblr. The goal was to test whether Middlebury could spark meaningful engagement and audience growth in less saturated spaces.
- Pinterest: By creating a board dedicated to AI-generated images, Andrew grew Middlebury’s Pinterest following by 50% in just one month. Pinterest’s DIY, aspirational vibe aligned well with the higher ed audience.
- LinkedIn: Posts featuring AI-generated images sparked alumni engagement, with users recognizing their own contributions and sharing them widely.
- Reddit: Andrew shared the content on subreddits focused on generative AI and Vermont, further diversifying the audience.
- Discord: A growing platform for Middlebury, Discord saw modest success as part of the campaign.
How Generative AI Works
Generative AI tools like MidJourney work by predicting what an image might look like based on a text prompt. The challenge is crafting detailed, descriptive prompts—what Andrew calls “prompt engineering.” For example, a simple prompt like “iceberg” might result in a generic image, but a more specific prompt like “a glowing iceberg floating in a dark ocean” yields a more evocative result.
Andrew likens the process to writing alt text for images, a skill that ensures visual content is accessible and descriptive. With practice, prompt engineering becomes an invaluable tool for creating visuals that resonate.
Results: Growth, Engagement, and Authenticity
The campaign produced tangible results across platforms:
- 53 Prompts Submitted: Over 50 descriptions of Middlebury were turned into AI-generated visuals.
- Increased Platform Engagement: A 50% growth in Pinterest followers, hundreds of clicks on the campaign’s bit.ly link, and dozens of LinkedIn comments from engaged users.
- Community Buy-In: Alumni and followers felt a deeper connection to the campaign, especially when their contributions were highlighted in posts.
By prioritizing community co-creation, Middlebury achieved more than just audience growth—it fostered genuine relationships and strengthened its brand authenticity.
Lessons Learned
1. AI Can Be a Tool for Engagement, Not Just a Gimmick
Andrew’s project went beyond simple engagement tactics. By involving the Middlebury community in the process, he turned AI-generated visuals into a storytelling tool that celebrated the college’s culture and values.
2. Go Where the Audiences Are—All of Them
While Instagram and Facebook remain dominant, platforms like Pinterest and Reddit offer untapped potential for higher ed marketers. The success of Andrew’s campaign proves that it’s worth experimenting with niche platforms.
3. Be Transparent About AI
One challenge Andrew faced was navigating public perception around generative AI. Concerns about copyright, ethics, and job displacement are valid, so it’s crucial to communicate how and why AI is being used in your campaigns.
4. Start Small and Build
Andrew’s initial goal was modest—30 prompts in 30 days without using Meta platforms. By starting small, he was able to test his ideas, refine his approach, and build a foundation for future campaigns.
5. Co-Create with Your Community
The real magic of Andrew’s campaign wasn’t the AI—it was the collaboration. Inviting your audience to participate in content creation builds trust, engagement, and a sense of ownership.
Why This Tactic Works
Andrew’s tactic works because it combines cutting-edge technology with timeless principles of engagement: listening to your audience, amplifying their voices, and creating content that resonates on a personal level. Generative AI isn’t just a trend—it’s a tool that, when used thoughtfully, can help higher ed institutions connect with their audiences in new and meaningful ways.
Ready to Try It?
If you’re inspired by Andrew’s story, here’s how to get started:
- Set a Goal: Decide what you want to achieve—whether it’s audience growth, engagement, or showcasing your institution’s culture.
- Ask for Input: Use a Google Form or survey to collect prompts from your community.
- Learn the Tools: Experiment with generative AI platforms like MidJourney or DALL-E, and practice crafting descriptive prompts.
- Choose Your Platforms: Diversify beyond Instagram and Facebook—try Pinterest, LinkedIn, or Reddit.
- Track Results: Use analytics tools to measure engagement, clicks, and audience growth.