Work Faster with AI: 3 Easy Prompting Techniques You Can Use in Any Workplace

Knowing how to prompt is quickly becoming an essential skill in any workplace, similar to what was happening several years ago with the Microsoft Office suite. When you did not know how to create tables in Excel or documents in Microsoft Word, you were basically almost unhirable in many corporate jobs.

And the same is happening now with AI.

Many companies have already integrated AI tools and platforms into their environment, and they expect you to be able to use them, or to at least understand the logic enough to learn and use them later. This is why prompting matters quite a lot in today's work environment.

Maybe you think: Why is that? Prompting itself does not seem like rocket science, right? You just open an AI tool like Claude or Microsoft Copilot, type in what you want, and get some result. That's it. What is there to learn about prompting?

Well, the thing is that when you prompt badly, you do not actually work faster or more efficiently. In fact, it can be quite the opposite. You get bad AI outputs, deal with hallucinations, correct generated text or images over and over again, and end up thinking: why is everyone so hyped about AI? Why should I use it, when I would be much quicker doing all of that by myself?

Now, obviously, AI has some limitations. This is why we need to verify AI results and correct them. Since we cannot expect everything that AI gives us to be the truth.

But many times, bad AI output is not about AI being stupid or hallucinating. In fact, it is about us writing a bad prompt, or simply not understanding how AI prompting for work is different from just typing a random request into a tool.

And that is why learning how to prompt AI and knowing a few practical AI prompting techniques actually makes a lot of sense.

Therefore, in this article I want to introduce 3 helpful and efficient prompting techniques that can help you in your daily work, based on the type of task you are trying to solve.

Now, before we jump into that, let me quickly stop by the actual prompt structure and how a well-written prompt should usually look.

Before we start: how should a prompt usually be structured?

When writing a prompt, it is helpful to imagine that you are not giving a command to a computer system, but that you are giving an instruction to a new colleague.

And if you had a new colleague sitting next to you, you probably would not just say: "Prepare the email" or "Create that report" and expect a perfect result.

You would explain what the task is about, who the output is for, what should be included, what should be avoided, and how the final result should look.

And AI needs a similar level of context.

A useful prompt usually includes a few basic things like: a clear task, enough context, some limitations, the audience, and the format or tone of voice you want.

For example, if you need only three sentences, say it. If AI should not invent data, tell it not to do this. If the output should be written for senior managers, candidates, customers, or your internal team, mention that as well.

Of course, you do not need to use all of these blocks every single time. Sometimes an easier prompt is enough. But in office work, especially when you need the output to be usable, the clearer you are, the less time you usually spend correcting the result afterwards.

Let's now move to the actual prompting techniques, and see how we can prompt differently based on different tasks and get better results from AI.

1st prompting technique — Iteration

The first prompting technique is iteration.

Iteration means that you refine the AI output step by step in one conversation, in one chat thread. This technique usually starts with a bit broader prompt, and then you keep adjusting the output with each next prompt until you get closer to what you actually need.

Let's say you start by asking AI to prepare an email for your team. So AI writes the first version.

With the next prompt, you ask it to make the email shorter. Then you ask it to change the tone. In the next prompt, you tell AI to adjust the email so it reflects the meeting you had last week. And after that, you can ask it to include some current company updates or remove parts that are not relevant anymore.

So with each prompt, you are refining the final output.

This technique is helpful when you have a rough idea of what you want, but you do not expect the first result to be perfect. And honestly, in many workplace situations, the first AI output will not be perfect anyway.

The goal here is to take that vague, general and generic first output and guide it further with next prompts.

So instead of trying to create the perfect prompt from the beginning, you can start with an average first prompt and then improve the result step by step.

2nd technique — Prompt Chaining

The second technique is prompt chaining.

Prompt chaining means that you split one complex task into several smaller prompts and steps. So instead of asking AI to do everything at once, you create a workflow and solve the task piece by piece.

Let's say you want to create a marketing campaign.

You could start by asking AI to help you gather insights about your target group. Then, in the same chat, you give it the next task and ask it to create an analysis from the data you provided. After that, you may ask it to suggest campaign messages. Then prepare social media posts. And later maybe use all of that to put together a presentation or campaign plan.

So you basically have one bigger task, but you are not trying to solve everything in one prompt.

This is useful because when you give AI one huge assignment with too many expectations, the result can easily become too general. The tool may understand the task only on the surface, skip important details, or give you an output that looks complete but does not really go deep enough.

Also, it is very difficult to define everything in one big prompt. You need to explain the task, context, limitations, audience, structure, and expected result (and when the task is complex, this can become quite messy).

So that is why it often works better to split the work into separate steps (prompts).

You create one piece, then another piece, then the next one, and at the end you can put everything together into the final output.

Prompt chaining is especially helpful for bigger tasks where you need some logic, research, analysis, writing, and final formatting. It gives you more control over the process, and you can check each step before moving to the next one.

3rd technique — Interview-Style Prompting

The third technique is interview-style prompting, and I think this is one of the most helpful ones.

Many times, we are dealing with complex tasks where we do not even know where to start. And if we do not know where to start, it is obviously very difficult to give AI a good prompt.

In this case, your prompt will probably be too generic, and the result will be too generic as well.

With interview-style prompting, you put yourself into the position of the interviewee, and AI becomes your interviewer.

You start by telling AI what the bigger task is, what you need to create or solve, but you do not need to give all the details immediately, because maybe you do not even know what details are needed.

Instead, you ask AI to interview you first.

So, for example, if we stay with the marketing campaign example - you can tell AI that you need to create a marketing campaign for a specific product, but you want it to ask you questions first, so you can define the full picture together before creating the final output.

AI can start by asking what the target market is. You answer for example Spain. Then, it can follow up with questions like: who the target group is, what the product does, how the company presents itself, what the campaign goal is, what channels you want to use, or what budget or limitations you have.

And after this back-and-forth conversation, AI has much more information to work with.

The useful thing here is that AI actually helps you identify what information is missing. It asks you for details that you may not have thought about at the beginning, and once the full picture is clearer, the final output is usually much better than if you tried to explain everything in one huge prompt from the start.

One small tip here: when you work with a bigger task, it can be quite tiring to write all the answers and long prompts manually. Therefore I would really recommend trying voice dictation or talking to AI, if the tool you use allows it. It often feels much more natural, and it is usually less time-consuming than trying to explain everything perfectly in writing.

How these techniques can help in recruitment and HR

And since I am a recruiter and consultant, let's put these three prompting techniques into the recruitment and HR perspective as well.

With iteration, where you keep adjusting the AI output with each prompt, this can be very helpful when for example creating job descriptions.

You can start by asking AI to prepare a general draft of a job description, and then you keep refining it. With each prompt you add more details, more specifications, more context about the team, the project, the seniority, or the tone of voice you want to use.

So the final job description is not created by one magical prompt. But it is created by several smaller adjustments until you are happy with the final result.

When it comes to prompt chaining, this can be helpful, for example, when doing market mapping or talent analysis.

Let's say you need to map talent availability in a specific region, maybe because your company is considering opening a new office there or hiring for a new role in that location.

Prompt chaining means you will split this task into several steps.

One step could be telling AI to gather information about the number of IT graduates in that location. Another step could be then identifying technical universities or relevant educational institutions in the area. Then you can tell AI to map companies hiring similar talent, then check how the role is commonly named on the local market, and later collect information about salary ranges or legal basics that may influence hiring.

At the end, you can take all of this information and let AI turn it into one market mapping document, PDF, or a presentation.

So here, instead of telling AI, "Create a talent analysis for this location," and expecting one perfect result, you create the analysis step by step.

And the third technique - interview-style prompting, can be very helpful when you want to design or redesign your hiring process.

For example, let's say your company needs to grow quickly and hire a larger number of people in a short period of time. You need the hiring process to be efficient, but maybe you are also not exactly sure what should change or where the biggest gaps are.

In this case, you can give AI the bigger task and explain the current situation of the company. Then you ask it to interview you on all the details it needs to know.

AI can ask about your current time to hire, how many people are involved in hiring, which locations you operate in, how your hiring process works now, where candidates usually drop out, or what roles are the hardest to fill.

Based on this information, AI can then help you identify gaps, suggest new workflows, and create a more structured hiring plan for the number of hires you need to reach in a given period of time. The point here is that you tell AI your goal and let it interview you for the needed details.

We can imagine AI being here an experienced consultant who comes to your office, gathers information on the task and situation, and then starts the execution without your further assistance. Whereas normally, we treat AI like a junior colleague - we give it strict commands ourselves and watch closely its every step (and then are often surprised that it did something wrong because we forgot to explain some important detail to it).

Final thoughts

Prompting is not only about writing one perfect sentence into an AI tool.

It is more about understanding how to give AI enough information, how to guide it, and how to choose the right approach based on the task you want to solve.

Iteration is useful when you want to improve an output step by step.

Prompt chaining is helpful when you need to split a bigger task into smaller parts.

And interview-style prompting works very well when the task is complex and you do not even know what information AI needs from you at the beginning.

The more you understand these techniques, the easier it becomes to use AI in a practical way, not only as something interesting to test, but as a tool that can actually make your work faster and easier.

And if you work in recruitment or HR and still feel unsure about AI, or you want to explore more practical use cases for using GenAI in hiring - feel free to check my AI in Recruitment and HR course. Now with the code AILEARN you can obtain the course for the lowest price possible of 9.99 EUR (valid until 27th August).

AI in Recruitment and HR course →

In this course, I cover not only AI basics and prompting, but I also show more than 30 practical use cases for tools like Copilot and ChatGPT, so you can get a better idea of how to use AI in recruitment and HR work.