How Core Messaging is Vital to Using AI Effectively

“Artificial intelligence is not a substitute for human wisdom but a tool to amplify it. Without a clear message, even the smartest algorithms can’t connect with your audience.”  — Unknown

Maybe the end of everything is coming, and we’ll all be reduced to batteries for our AI overlords. But until then, we press on in our mission of using AI as a tool for good, especially in the nonprofit space.

At Liminal, one of the questions we regularly ask ourselves is this:

How can nonprofits use tools like ChatGPT Gemini, and Cluade responsibly, ethically, and effectively without losing their voice or creating more boring or generic, obviously AI-generated content?

We’ve seen a range of responses to AI adoption across the nonprofit sector.

Here are four of the most common:

1. Avoidance: Some people are afraid of AI, think it’s a fad, or just don’t understand it, so they completely ignore it.
2. Apathy: Others don’t resist it, but they don’t engage with it either. It’s just not on their radar, and they are not utilizing it at all.
3. Uncritical Overuse: The organizations are all in, using AI for everything, but without clear guidance or strategy. What they gain in speed, they lose in clarity and voice.
4. Strategic Implementation: These are the teams that view AI as a tool, and the results they get are only as good as the inputs they provide. They understand that AI won’t think for them but can be a partner or tool in their work.

That last group of leaders is where we should land, but uncritically jumping into the AI deep-end is not the answer.

If you want AI to produce compelling content, you must first train it to understand who you are, what you do, why you do it, who you serve, etc.

We work with nonprofits to develop a comprehensive Core Messaging Stack. This strategic document becomes the foundation for everything: website, cases for support, grant writing, fundraising campaigns, internal comms, onboarding. It also happens to be incredibly helpful for when a board member has to explain what you do at a dinner party.

As a nonprofit leader, your core messaging content should inform everything you do with AI. You should rarely open ChatGPT (or your language model of choice) and start creating content or asking for strategic advice and feedback if it doesn’t have anything to work with.

Of course, unless you are using it to ask about the best coffee-to-water ratio for your pour-over. In that case, you can just open the app and fire away.

Informing your AI language model is essential to drafting unique, compelling, and engaging content.

Hopefully, you noticed we said “drafting” content, not “writing content that you can uncritically copy and paste and drop into a marketing or campaign email.”

Most people have gotten extremely good at noticing AI-generated content’s cadence, writing style, and punctuation.

If your goal is to generate generic content that people ignore, by all means, open the app and have at it.

But if your goal is to utilize AI to help process new ideas, test out theories, and draft content that is informed and relevant, you need to inform AI with content and context about your organization, mission, target audience, and voice, etc.

You can do this by training it with a Core Messaging Stack.

Your Core Messaging Stack is a foundational document that outlines your organization’s mission, values, personality, tone, key messages, audience insights, and common talking points. When you inform AI with this content level, you create a training set.

Without foundational messaging to work with, AI will fill in the gaps with generic fluff. Training AI with a Core Messaging Stack is like giving AI a map. Without it, it’s wandering around, grabbing the latest uniformed jargon and generic copywriting.

But when it is trained with your content, it is able to navigate with purpose.

Here’s why it matters.

If you’ve used AI tools like ChatGPT, you know how easy it is to generate content that sounds polished but also feels “off.”

That’s because these tools are brilliant at structuring language, but they don’t know your context, audience, or the complex work your team is doing.

Again, that’s why you have to train it.

So what’s next?

If you’re exploring ways to make AI work for your nonprofit, start here:

• Build (or refine) your Core Messaging Stack (We can help with this!) A core messaging stack should codify (at least) the following topics and statements: Why Statement, Vision Statement, Mission Statement, Core Values, Tagline, Brand Voice & Personality, Problem & Solution Statements, Programs/Services Descriptions, Central Characters in Your Story, Positioning, Target Audience, Calls to Action, Narrative (About Us), and an Elevator Pitch. We know its a lot, but nailing this content is incredibly important for your messaging as a whole, including utilizing AI.
• Teach AI tools using that messaging. If you use ChatGPT, creating custom GPTs is a great way to start.
• Train your team (and learn from those already using AI effectively) on how to utilize it in your work. Some on your team are already using AI; learn from them. Those who aren’t using it will need to be trained. Either way, you must be intentional about how your organization utilizes AI.
• Use AI for outlines and drafts. When do you use it for outlines and drafts, edit them ruthlessly and make sure that your voice is coming through in the final results.

AI overlords are not going to replace your work, at least not yet. However, those who do not utilize AI to enhance and speed up their work will be replaced by those who do, and those who use AI uncritically might lose their jobs.

Concerned, Confused, or Curious about how to use AI? We’re here to help. Developing a full Core Messaging Stack for Nonprofits and Social Impact organizations is our specialty — Learn More.