In recent decades, AI has experienced significant development, resulting in a series of innovations and updates that bring substantial benefits. 

One of the widely used AI-based technologies is generative AI. This chatbot technology enables users to generate outputs in the form of text, images, and even videos based on prompt instructions.

While many of us actively access this technology and benefit from it, we often find that the outputs do not always meet our expectations.

This highlights the importance of correct and precise prompts.

So, what should we do to maximize the performance of generative AI and achieve outputs that meet our expectations? Read on to find out!

Generative AI Prompts

Before diving into strategies and tips for writing prompts, let's first understand what prompts are.

Prompts are the process of forming an instruction or question directed at generative AI to produce the desired output.

Why is this important? Without the right prompts, we won’t get the maximum results from generative AI. To optimize the outputs, here are some tips you can follow:

• Act as If

Generative AI can generate various forms of information and data based on the latest updates in technology.

To get specific outputs, we need this technology to think according to a certain role or profession to ensure the output’s perspective and tone meet our needs.

For example, if we want to create an editorial plan for social media, we can add a role to the prompt, such as "act as a social media specialist and create an editorial plan for social media..."

This way, the output we get will align with our needs in the scope of social media marketing.

• Define Objective

The next step is to define what we need.

Generative AI can simplify and speed up our work in producing information. We can provide various instructions that will be transformed into insights, images, or videos.

For example, if we need a series of sentences for a webinar invitation email, the prompt can be "act as a copywriter and create a series of sentences for a webinar invitation email..."

Generative AI will produce output according to the instructions given.

• Specificity is Key

The more specific the prompt, the more specific the result.

This is crucial when we want to incorporate a certain tone or image for our brand. In such cases, we need an output that stands out and can be personalized to match the brand image we want to build.

Several aspects can be used to write specific prompts:

• Platform

Which platform will the output be used on? For example, Instagram, email, website, etc. Including this aspect in the prompt will ensure the output matches the tone of that platform.

• Audience

Sometimes, we need wording that is targeted to a specific group, such as women, seniors, workers, etc.

• Length

The length of the wording can also be tailored to our needs or objectives. For instance, if we need wording for a caption, the length can be adjusted to between 80–100 characters.

• Tone/Style

In marketing, engaging the target audience is crucial. Whether the wording should have a tone of excitement, persuasion, or attractiveness can be specified in the prompt.

• Leverage Visual

When we want the audience to visualize a situation, scent, or image, we can instruct the prompts to generate sentences that achieve this.

For example, if we want the audience to imagine the beautiful scenery of Japan with cherry blossoms and koi fish in the ponds, generative AI can produce the output to match this need.

• Do/Don’t

For more specific outputs, include details on what should or shouldn’t be in the results. For example, if we want certain data or statistics included, we can specify this in the prompt.

• Example

You can attach a sample wording you have as a reference for generative AI to produce similar outputs. This method is more practical than manually inputting multiple items.

• Refine the Result

Once you have the required result, the final step is to validate the output generated by generative AI. Ensure the result is valid, precise, and meets your needs.

Conclusion

The steps above can be applied when writing generative AI prompts. However, these steps are not rigid. You can discover your methods and strategies based on your experiences and trials. Each element can be combined to achieve optimal results.