Re-thinking AI outputs with prompt strategy

Did anyone notice how, in 2025, “AI-generated” quietly became shorthand for “low quality”? I’ve been thinking a lot about that lately… and why. It’s not like the technology suddenly got worse — I think with everyone using it, for everything, our expectations got higher and our outputs couldn’t keep up. We started treating AI like a vending machine instead of a creative collaborator that needs direction, and when left to its own devices, we’ve now seen just what it can (and can’t) do.

The challenge with AI is no longer about adoption — we’ve moved well beyond that. The challenge now is about moving beyond rushed, robotic, interchangeable outputs and using AI in ways that actually strengthen creative thinking instead of flattening it.

Sometimes seeing is believing, so if you’re willing, give this a try. Open your favorite LLM (ChatGPT, Gemini, Claude, whatever you use) and ask it to generate a 100-word blog post about cats. No other instructions. And save your output somewhere.

Now ask the same thing again — but this time, include some prompt strategy in your request:

• Who the target audience is

• What the voice and tone should be (friendly, serious, conversational, educational, etc.)

• What you want the reader to think, feel, and do after reading

Compare the two outputs… Do you see what I see? 👀

Without a prompt strategy, most large language models fall back on the same familiar patterns — structures and phrasing they’ve learned over time — and just generate more of the same. If we want to find real ways to differentiate those outputs, and get to a stronger first draft faster, prompt strategy is going to become a core creative skill. From there, the real work begins: refining, shaping, and bringing the content to life.

Don’t worry! This isn’t another post about a course you need to take, a text book you need to read… you probably already have these skills! You just might not yet have thought about how to bring them to the AI table. 

In my experience, a strong prompt strategy relies on the fundamentals that us content folks already use in our day today, including:

• A clear point of view

• Deep audience understanding

• Brand voice constraints

• Context, references, and boundaries

• Iteration and critique

Like I said — none of this is new. What’s new is where and how we apply it, when we’re working with AI.

As content practitioners, we know what good looks like before we start. We can sense what we want an audience to feel. We can anticipate the conversation we want to spark. AI isn’t going to replace that intuition — but it depends on it to create interesting and valuable outputs. When we feed our expertise and instincts into our prompts, we don’t lose our craft — we start to reclaim it.

Prompt strategy isn’t about clever commands — that’s prompt engineering and that’s a whole different conversation. Prompt strategy is about weaving human insight with creative intent. Without it, we’ll just keep getting more of the same. And no one wants that anymore.