Another excellent article on AI from Vittorio Neri, Head of Marketing Research, Planning & Technology at Roland DG EMEA.
For professionals who want to go beyond the basics, whether you work in digital printing, visual communication or product personalisation, AI can become not just useful but strategic.
Not everything is a prompt. Sometimes you need a strategy.
If you’ve explored tools like ChatGPT, Gemini, Claude, Perplexity or Grok, you know they can be incredibly helpful. But at some point, you hit a limit.
You realise it’s not just about asking better questions. It’s about asking the right ones, at the right time, with the right goal.
That’s where strategy comes in. The difference between good AI and great AI often depends not on the tool (they are overwhelmingly powerful today) but on the method.
When is a prompt not enough?
There are three points to consider. You might get:
- Outputs that sound right, but lack substance
- Ideas that look good, but don’t fit your business
- Texts that are polished, but generic
The problem? It’s not the AI. It’s the absence of context, of structure, of clarity. This happens quite often, especially after the initial phase, when you start to crave more.
Think like a strategist, not just a prompter
Before opening the chatbot, ask yourself:
- What’s my objective? (Inspire, decide, explain?)
- Who is this for? (Client, partner, internal use?)
- What does success look like? (A draft, a concept, a decision?)
Then design your interaction accordingly. Don’t ask for “something”. Frame a path.
Instead of: “Write a post about our new fabric printer.” Try: “Draft a LinkedIn post to announce our new (fabric/DTF) printer to B2B clients in the hospitality industry for on-demand deliveries. Tone: fresh but professional. Goal: start conversations explaining what services we can now offer with our new equipment.”
Use frameworks and templates
Strategic prompts aren’t necessarily long, but they’re structured. Use repeatable formats:
- Goal → Audience → Message → Style
- Problem → Insight → Proposal → Next step
Create templates that reflect your workflow:
- “Generate a three-post series to support a product launch in [industry]. Include: teaser, product intro, behind-the-scenes.”
- “Build a customer profile based on this company’s website and LinkedIn. Then write a cold email offer tailored to them.”
Remember, the more context you can offer to your AI tool, the better the results will be.
AI Insights & Tricks
Here are a few practical tips that make a real difference when working with AI:
- Stack your prompts. Instead of asking one big question, break it into steps. Ask first for ideas, then narrow down, then reframe. Iterate as much as you can. You can also open two windows and, in one window, create the prompt, while in the other, test the results. Then, copy and paste the results into the window where you are building the prompt, asking the AI to verify if the answer matches the instructions.
- Give it your voice. Share an example of your tone or writing style, then ask AI to match it. You can share an article from your blog or website if you already have a specific tone of voice. Otherwise, you can select articles from authors you like and ask the AI to follow the same patterns.
- Name your frameworks. Tell the AI you’re using a model like “PAS” (Problem-Agitate-Solve) or “AIDA” (Attention-Interest-Desire-Action). These are classic copywriting and messaging structures. PAS highlights a problem, stirs emotion, then offers a solution; AIDA captures attention, builds interest, creates desire, and drives action. Using them helps the AI organise your content more effectively.
- Use delimiters for clarity. When providing lengthy instructions or pasting reference text, use tags like <<<this is the source>>> to maintain a tight context. In this case, providing an example of the expected answer structure can make a difference.
- Prompt with purpose. Add a sentence like “The goal is to sound confident but friendly” or “This is going to be shared on LinkedIn, so keep it conversational.”
The more you think like a creative director or strategist when prompting, the more useful and aligned the responses will be.
Conversational vs Reasoning to choose the right brain
Not all AI models think the same way.
Some are brilliant at natural language and creativity. Others focus more on analytical reasoning, logic and structured thought.
Here’s how to make the most of them:
- Use conversational models for copywriting, naming, storytelling, tone of voice, and social content, as well as quick analysis.
- Use reasoning-oriented models for decision trees, problem-solving, comparing options, analysing documents, and extracting insights.
The trick? Start your prompt by stating the nature of the task.
“This is a reasoning task. Help me compare three service models and decide which is more scalable for a small print business.”
Matching the type of brain to the type of task is half of the success.
Quick tip. The lines between these model types are evolving fast. In the near future, we may see models that blend reasoning and conversation more seamlessly. Keep an eye on new releases and try different tools for different types of thinking.
Mini-experiment
“Act as a strategist for a small print studio that wants to reposition its large-format services for events and fairs. Suggest two positioning concepts, each with key messages and a campaign idea.”
Then follow up:
“Turn concept 1 into a landing page text.” “Create a short sales pitch for concept 2, focused on price/quality balance.”
Closing thoughts
Prompts are just the beginning. AI becomes a true creative and operational partner when paired with strategy, clarity, and iteration. This new series is about that next step. Not just tools, but mindset, method and mastery. Let’s start better, together. Feel free to reach out to me if you’d like to share your experience.
