Common mistakes when writing prompts
The ten most common prompt-writing mistakes and how to avoid them: from missing context to blindly trusting the answer.
These are the mistakes we see most often and how to fix them.
1. Being too vague
“Help me with my website” forces the AI to guess. Specify which part, which problem and what result you expect.
2. Giving no context
The AI does not know your business, audience or situation. Whatever you do not tell it, it assumes.
3. Asking for too much at once
A single prompt with ten tasks produces ten mediocre results. Split it into steps and review each one.
4. Not specifying the format
If you need a table to paste into a spreadsheet, ask for it. Otherwise you will get paragraphs.
5. Contradictory instructions
“Very detailed” and “very short” at the same time. Reread the prompt before sending it.
6. Accepting the first answer
The first result is a draft. Ask for specific improvements.
7. Not verifying data
Models can invent figures, quotes, links or references with complete confidence. Always check what matters against reliable sources.
8. Sharing sensitive information
Do not paste passwords, customers' personal data, medical records or confidential company information into tools that are not authorized for it.
9. Using AI for decisions that need a professional
AI can help you understand a legal, medical or financial topic and prepare questions, but it does not replace a licensed professional.
10. Not saving the prompts that work
When a prompt works, save it as a template with variables. It saves time and gives consistent results. You can also save the library prompts you use most to your Favorites.