Introduction
How do you write an effective AI prompt without sounding like a robot? That question guided the latest Marketing Meetup session with Anna Carina Berkman, also known as Architect Anna, an AI trainer and consultant who helps people get better results from AI tools.
Anna believes that good prompting is simply good communication. In this session, she shared how to master AI prompts in 2025 using her COCO Framework and a practical technique she calls “lazy prompting.”
Here is a summary of the main insights from the latest discussion in our AI season. We’ve used an AI helper to write this with us, so please excuse any pesky errors 🙂
Table of Contents
- Introduction
- What an AI Prompt Really Means
- The COCO Framework: A Simple Way to Write Better AI Prompts
- Lazy Prompting: How to Prompt Without Prompting
- Building an AI Prompt Library
- Personalising Your AI Tools
- Why Critical Thinking Is the Ultimate AI Skill
- Conclusion
What an AI Prompt Really Means
An AI prompt is the instruction you give to an artificial intelligence tool such as ChatGPT, Claude or Gemini. The clearer the prompt, the better the output.
Anna explained that prompting has not changed much since 2022. Whether you are creating content, building presentations or analysing data, the core principle remains: describe your goal clearly and provide enough context for the AI to understand what you need.
She compared a chatbot to a car: the text box is the steering wheel, but the large language model (LLM) behind it is the engine. Your AI prompt is how you steer.
The COCO Framework: A Simple Way to Write Better AI Prompts
Anna’s COCO Framework makes creating a strong AI prompt simple. It stands for:
Character – Objective – Context – Output
- Character: Give the AI a role, such as “Act as a marketing strategist.”
- Objective: Define your goal clearly, for example, “Help me plan a presentation.”
- Context: Add background, tone, audience and examples.
- Output: Specify what you want in return, such as “Write a three-paragraph summary with a call to action.”
Following COCO makes any AI prompt clearer, whether you are writing blog posts, sales emails or campaign scripts.
Lazy Prompting: How to Prompt Without Prompting
Anna’s second technique, which she calls lazy prompting, helps users get better results with less effort.
Instead of crafting a long, complex AI prompt, describe your goal and ask the chatbot to ask you one clarifying question at a time. This back-and-forth approach allows the AI to collect context gradually and produce a more accurate result.
For example:
“I need to create a presentation for tomorrow’s marketing event. Ask me one question at a time to gather everything you need to help me finish it.”
This conversational method reduces confusion and builds collaboration between human and machine.
Building an AI Prompt Library
Once you’ve created a few good prompts, save them in a prompt library. Anna recommends keeping them in a shared document, such as Google Docs or Notion, so your whole team can access them.
Each AI prompt should include tone, structure and examples of good outputs. Review your prompt library every few months to keep it relevant as tools evolve.
If you regularly work on complex projects, consider building custom GPTs, specialised assistants trained on your content and tone. Otherwise, a simple reusable prompt list works perfectly.
Personalising Your AI Tools
Anna also highlighted how important it is to personalise your chatbot. Tell it about you, your writing preferences and your style. For example, “Write in British English, avoid em dashes and keep the tone conversational.”
Switch off model improvement in your settings to protect your data, and customise the tone to match your personality. This helps every AI prompt feel more natural and aligned with your voice.
Why Critical Thinking Is the Ultimate AI Skill
Even the best AI prompts need human review. Anna reminded everyone that AI systems are trained on biased and incomplete data. They can hallucinate facts or make assumptions based on Western or internet-heavy sources.
Critical thinking is the ultimate skill for anyone using AI tools. Always check facts, verify sources and add your own expertise. The difference between a beginner and an expert is not in how many prompts they write but in how well they analyse what AI produces.
Conclusion
Writing a strong AI prompt is not about technical tricks. It’s about clarity, creativity and context. Anna’s COCO Framework gives marketers and creators a simple way to master prompting, while lazy prompting saves time and keeps collaboration human.
Her advice was simple: treat AI as your 24/7 success partner. When you give it direction, it helps you achieve your goals faster.
Prompting, when done well, is not just about teaching machines to think. It’s about learning to think better ourselves.