Many people make mistakes when prompting because they are too vague or unstructured. The solution: A Clear structure for AI prompts use! In this article, we'll show you how to use this technique for better AI prompts uses a concrete example.
You have probably already experienced that an AI like ChatGPT does not deliver exactly what you expected. The answer is too superficial, misses the point or is incomplete. Sometimes you get a long, unstructured explanation, even though you actually wanted a compact list. Other times, crucial details are missing or the AI adds information that is not relevant.
The problem is often not with the AI itself, but with the way you formulate your query. AI models have millions of ways to solve a task. If you don't ask your question precisely, the model itself decides which information is relevant - and that doesn't always correspond to what you had in mind.
The good news is: You can specifically control how the AI responds. OpenAI President Greg Brockman, in connection with ChatGPT o1 emphasises that a well thought-out prompt structure is crucial for more precise and relevant results. In a repost he emphasised the importance of clearly formulated queries in order to make the most of the strengths of this model. In this article, we show you how you can optimise your prompts with simple adjustments and thus significantly improve the quality of AI responses.
ChatGPT o1 demands a new way to prompt - Greg Brockman's insights
How the optimal AI prompt works
For AI models to deliver more precise results, they need to understand exactly what is expected of them. A structured approach helps to significantly improve the quality of the answers. There are four central building blocks:
- A clear target
- A defined output format
- Restrictions for unwanted content
- Additional context for better categorisation
These four building blocks ensure that the AI provides exactly the answers that are needed instead of choosing one of many possible solutions at random, and you will never again be annoyed that the AI does not understand what you actually expect.
1. clear target - tell the AI exactly what you want
A common mistake is that requests are formulated too generally. For example, if you ask for a travel recommendation and only write "Recommend a good travel destination." gives the AI far too much room for manoeuvre. The answer could be a popular tourist hotspot, a remote location or something completely inappropriate.
A more precise formulation would be better: "I am looking for the three best weekend trips in Europe that offer special experiences and are less crowded." This variant is better because the AI immediately recognises that several options are required instead of just giving a recommendation. The default "special experiences" restricts the selection to places with a unique offer, while the "less crowded" restriction prevents typical tourist hotspots such as Paris or Rome from being suggested.
The more precisely the goal is formulated, the more relevant and usable the AI result will be.
2. desired output format of the prompt - define the structure of the response
AI models tend to provide answers in long continuous texts, although a clearly structured list would often be more helpful. A vague query such as "Which destinations can you recommend for a short trip?" often results in the AI generating an unstructured paragraph describing various places without a clear separation.
An optimised variant would be to give the AI more precise instructions: "The following information should be provided for each destination: Name of the location, distance from Berlin, best time to travel, recommended activities and highlights, and the average cost of accommodation and food." This ensures that the answer not only contains relevant information, but is also directly comparable. Unimportant details are avoided and the output is immediately usable.
If an even clearer structure is required, you can explicitly request a tabular representation by adding the following to the prompt: "Output the answer as a table." This makes the output even clearer and easier to analyse.
3. what AI should avoid - reduce the risk of errors
A common problem with AI-generated answers is that they provide inappropriate or irrelevant suggestions without clear restrictions. A vague query such as "Which cities are suitable for a short trip?" gives the AI too much leeway, meaning that it may recommend overcrowded tourist locations or destinations that are difficult to reach.
A better solution is to give the AI precise restrictions, such as: "Please don't suggest overcrowded tourist hotspots like Paris or Rome. Prefer destinations with good public transport connections and unique experiences." In this way, the AI takes into account which places should be avoided, provides more suitable suggestions and reduces the risk of queries or corrections.
For more complex queries in particular, it is crucial to not only formulate what the AI should do, but also what it should avoid. This makes the results more targeted and better tailored to the actual needs.
4. provide context - give the AI relevant additional information
Although AI models have an enormous amount of knowledge, they cannot automatically recognise which individual factors are decisive for a query. An overly general prompt such as "Recommend me a destination for the weekend" does not give the AI any clues about personal preferences or travel preferences. The answer could therefore include anything from a city break to a beach holiday - without taking the user's actual needs into account.
To get more precise and relevant suggestions, the prompt should contain more context. A better variant would be: "I'm travelling with my partner, we like nature, good food and authentic cultural experiences. The trip should be relaxed and not include an overloaded sightseeing programme." With this additional information, the AI can adapt the recommendations much better instead of guessing in the dark. The results are therefore much more precise and immediately applicable, without the need for many subsequent adjustments. The following applies: The context should be as detailed as necessary, but not unnecessarily long, so that the AI can correctly prioritise the relevant information.
Conclusion: Why a clear prompt structure is crucial
The way in which you formulate an AI prompt has a direct influence on the quality of the response. Vague and unstructured prompts often lead to inaccurate or unhelpful results, while a clear structure ensures that the AI delivers exactly what you need.
By optimising your prompts in a targeted manner - with a clear objectives, a defined output format, restrictions on unwanted content and additional context - you can make sure that you receive more precise, relevant and immediately usable answers.
If you Support with the optimisation of your AI workflows we will be happy to help you. Would you like to further optimise your AI prompts or automations? We support you with customised strategies - contact us for an individual consultation!
KI Consultancy in Cologne
Your personalContact us
We look forward to every enquiry and will respond as quickly as possible.
KI Consultancy in Cologne - Contact form
KI counselling in Cologne - your personal contact
Good Business relations begin in person.
Contact us us with pleasure per Mail or Telephone, and we agree one personal Date.
German 
English