How to use ChatGPT more effectively at work

Why context determines the answer

The same request can generate completely different texts depending on who receives them. Without clear context, the model cannot determine whether a summary will be read by a director, a technical team, or presented to clients.

When a request includes information about the task, target audience, and expected outcome, the system can reduce ambiguity and guide the response more accurately.

The question that precedes the request

Before sending a request to ChatGPT, it’s worth mentally answering three key questions:

• Who will use or read the answer?
• What is the purpose of this answer?
• How should the result be presented?

OpenAI itself recommends specifying the audience, objective, and desired format. This guidance applies to tasks such as document writing, data analysis, and preparing professional materials.

How to apply this technique in real-life situations

Imagine a marketing analyst who needs to evaluate the results of a campaign. Instead of simply typing “analyze this data,” they can formulate a more detailed request:

“Analyze the data from a digital marketing campaign. The presentation audience is the board of directors. Identify the three main results, possible problems, and two recommendations. Present in short bullet points.”

The request provides enough context for the model to understand not only the task but also how the information will be used. The same principle can be applied to emails, reports, research, presentations, planning, and document reviews.

There is no perfect prompt

Experts caution that the initial request should not be considered final. If the response is inadequate, the user can continue the conversation and adjust what needs to be changed.

Requests such as “reduce by half,” “use more technical language,” “consider that the audience consists of executives,” or “convert it into a table” serve as successive refinements.

OpenAI advocates for this very iterative process, where the user evaluates the response and adds context or modifies instructions as needed.

Instead of searching for a complex formula for every interaction, professionals can start with a simple question: who needs this answer and why? From there, it becomes easier to define the necessary context, format, and level of detail for the task.