Prompting the AI Co-Pilot in RGE Studio

Beefree App is now RGE Studio!
The email design suite you already know and love just leveled up. We’re working hard to update our help center, but you still may see some references to the old product name.

Overview

AI tools help you accomplish more tasks more efficiently. To use AI tools effectively, you need to write a good prompt. However, writing an effective prompt can be more difficult than it sounds. We’re happy to share our suggestions with you!

This guidance is for the AI Co-Pilot in RGE Studio. You can also apply it to other AI tools. The AI Co-Pilot is currently in beta. To learn more about it, read our AI Co-Pilot article.

You can connect RGE Studio to your own AI assistant with our MCP server. The RGE Studio prompt library has ready-to-use prompts for that setup.

Why the prompt matters

An AI tool's output is only as effective as its prompt. This is because AI reads an entire webpage in a matter of seconds. It looks for content that matches your prompt. Then it repackages that content the way it thinks you want it. Without guidelines, it can make assumptions about how you want this to happen.

Your prompt tells the AI tool exactly what to do and what to avoid. The more specific you are, the better the output will be. If you’re not happy with the output, try improving your prompt.

Prompting the AI Co-Pilot

Prompting the RGE Studio AI Co-Pilot involves two components:

1. Create a written prompt. This is always required.
2. Set up your workspace styles in RGE Studio. This is optional, but recommended.

You need to give the AI Co-Pilot permission to use your workspace styles. To do this, click the apply styles button below the prompt field. It includes your workspace name. This action generates a .txt file containing the styles from your workspace settings. The generated file is automatically applied to your next prompt.

Your written prompt should describe the content and layout you want to create. The AI Co-Pilot will use your applied workspace styles during content generation. Alternatively, you can describe the styles you want to use directly in your prompt.

Below, we’ll discuss the written prompt and share an example.

The Prompt Formula

After years of testing AI tools, we created a formula. It works in almost every situation. There are three parts of this formula:

1. Context
2. Intended Output 
3. Direct Question

Here’s what to include in each part of your prompt.

Context 

Context provides the AI with background information about your prompt. Approach this as if you’re briefing a team member on a project. This element is important because it improves the quality of the output.

The context part of your prompt can include the following:

1. A high-level description of what you are trying to accomplish.
2. Any relevant background information. This may include general information about your brand styles and tone of voice.
3. An explanation of why you are trying to accomplish this.
4. A description of who your intended audience is.
5. If applicable, what you’ve already done or tried to do.

Depending on your use case, you can include only some of this information. Even a simple prompt should include context for the why and the who.

Intended Output

This section of a prompt is generally straightforward. However, you can also use it to define what the output should look like. This element is important because it helps the output meet your unique requirements.

The intended output part of your prompt can include the following:

1. A detailed description of what you’re trying to accomplish.
2. Tone and style guidelines for the output to follow.
3. Details or elements the output should or should not include.
4. If applicable, a request for multiple options or outputs to choose from.

The detailed description of your intended output is what’s most important here. It builds off the context you set in the first part of your prompt.

All other elements are optional. Use them based on your use case. In RGE Studio, you can add style guidelines by applying workspace styles to your prompt.

Direct Question

Some AI tools can hallucinate. This means they give answers that are false or made up. A direct question at the end of your prompt can help avoid this issue.

Consider this the AI’s call to action. This element is important because it keeps the AI tool on-task.

Example

Here’s a brief example of a prompt that uses this formula with our AI Co-Pilot:

[context]: "I'm designing an email to inform existing customers about an upcoming promotion. The promotion offers a discount on my clothing brand's summer collection. There is a limited stock and the promotion will only last as long as we have available items. We are making room for our upcoming fall collection."

The prompt starts with a very high-level description of what we’re trying to do. It also describes the audience (existing customers). It explains why we’re creating the content (making room for new inventory). Finally, it shares how long the promotion will be available.

[intended output]: "The email will be sent to a list of existing customers who haven't made a purchase in 6 months. It needs to encourage people to act fast without being too pushy. We'd like to highlight 3 clothing items that we have a high inventory of. The tone should be friendly and conversational."

This section expands on the who to provide clarity on the intended output. It describes what the designer is hoping to accomplish with the email. It defines some content that the designer wants to include. It also provides guidelines about the tone of the copy.

[direct question]: "Can you help me create an email for this promotion?"

The prompt ends on a direct question for the AI to act on.

Another Example

The prompt in this article is a basic illustration of our prompt formula. Every prompt is different.

We've attached another prompt example to this article. This prompt uses a template style. You can download it by clicking the link below. Feel free to repurpose it for your own needs!