Visual content is one of the most powerful marketing assets you have — photos, images, videos.

Do you know over 80% of what humans learn is through vision?

But if you’re like me and not good at design or art, that’s not encouraging.

Thankfully generative AI levels the playing field!

Imagination is literally the only limitation now.

In this post, I want to share some resources to help you level up your generative AI game—how to create images, find images, improve images, and even get an AI image of yourself.

Hopefully this will be as magical as watching a UFO land on earth 👽️ 

💡 Strategy: Steal The Images You Love

Pablo Picasso offers great advice when it comes to starting your generative AI art game: “Good artists borrow, great artists steal.”

Remember, I’m not artistic. However, I do have my own taste: I know what I want, and what I like.

Quick note: I use Midjourney, $10/mo. Ideogram is a new option that is free, and you don’t need to use Discord like with Midjourney (oh, and it renders text better than Midjourney).

Anyway, there are 3 ways that I get to the images that I like:

Strategy #1: Use AI Image Search Engines

If you know what you are looking for, use a search engine for generative AI images. This is helpful because it showcases the different styles you can come up with.

PromptHero is one that I have bookmarked.

For example, say you are a backpack brand and want to create an image instead of using your own or stock images.

You can go to PromptHero, and see a bunch of “back pack” related prompts and images.

This one is a cool example, in a “knolling” style that showcases how much a bag can hold.

And the great thing is that all of the prompts are included, so you can literally reproduce this on your own (own).

I tried in Midjourney with the exact same prompt, didn’t turn out as well, but you get the idea….

Strategy #2: Reverse Engineer an Image that you like

Midjourney has a great feature that lets you upload an image and it spits back a text description of the image.

So if you see any image you like, here’s what you do:

1. Save the image

2. Use the “/describe” feature in Midjourney.

3. Upload the image to Midjourney.

4. See how Midjourney interprets the image

5. Try the prompt in Midjourney

Here is how Midjourney describes the same knolling backpack example:

And when I tried to recreate with that exact prompt….

Not exactly as polished, but it’s a starting point!

Strategy #3: Use Prompt Guides to Create Better Prompts

The more details and descriptions you provide, the better the AI images that are created. The only downside is that you don’t know what you don’t know.

What kinds of prompts and parameters can you provide?

That’s where a prompt builder like PromptFolder can be helpful.

If you just have the seed of an idea, it can help you add the spicy details that make the image come alive, including the style, lighting, camera, artist, aspect ratio, and much more. Play with these settings, create an image. And iterate from there.

📚Use Case: Use AI Images Across Platforms

Now let’s get into the actionable use cases. You can get great mileage from these images in the following:

• Social media posts: Generate endless unique images that capture attention and align with messaging.

• Blog headers: Create colorful, conceptual images that relate to post topics and draw in readers.

• Presentation slides: Craft perfectly-worded slides that clearly explain key ideas and talking points.

• Email campaigns: Write compelling subject lines and copy that entice readers to open and engage emails.

• Ad creatives: Prompt for visual ad concepts and images that connect with target audiences.

• YouTube thumbnails: Design eye-catching custom thumbnails that represent video content and inspire clicks.

The key is using generative AI strategically across marketing initiatives to create captivating, optimized visuals and copy.

Brands like Nestle and Oreo use generative AI for their campaigns, saving 10-20x the cost!

Here’s a cool use case:

Cadbury used AI-powered ad templates for small businesses in India to have a global movie star, Shaah Rukh Khan, look like he is creating a personalized ad for them.

More than 130,000 ads were created with generative AI!

💪 Prompt & Practice Ideas

Generative AI for images isn’t only about a prompt, but learning how to understand what you like.

Here are some ideas to get started as you create your image:

1. Start by brainstorming your marketing goals. What do you want to achieve with your campaign? Once you know your goals, you can start to think about how Midjourney can help you to achieve them.

2. Identify your target audience. Who are you trying to reach with your campaign? Once you know your target audience, you can start to tailor your prompts to their interests and needs.

3. Write specific and descriptive prompts. The more specific and descriptive your prompts are, the better the results you'll get. For example, instead of prompting "generate a cool image," you could prompt "generate a photorealistic image of a new smartphone with a sleek design and cutting-edge features."

4. Experiment with different settings and parameters. Midjourney offers a variety of settings and parameters that you can adjust to get the best results. For example, you can adjust the style of the image, the level of detail, and the mood.

5. Edit and refine the generated content. Generative AI tools can generate some great content, but it's often necessary to edit and refine the results to ensure that they are perfect for your marketing campaigns. For example, you may need to crop the image, adjust the colors, or add text.

💥 Experiment: Try Different Image Styles

One of the best parts of working with AI image generation tools like Midjourney is the ability to experiment iteratively.

Testing different prompts and parameters allows you to refine your skills and get better results over time.

This is where PromptFolder’s tool is super helpful!

A few ideas to play with:

1. Vary the prompt wording: Reword prompts with different adjectives, descriptions, and ordering to see their impact on the output.

2. Modify the image style: Switch between realistic, abstract, and specific art styles to find what works best.

3. Include creative concepts or metaphors: For example “a book opening a portal to another world” or “personification of a startup’s brand essence.”

4. Blend multiple ideas together: You can prompt Midjourney to combine disparate concepts into a cohesive image. More on this in the future.

Continually trying new prompt approaches pushes your skills forward.

Pay attention to what wording and parameters produce your preferred visual style. Refine these over time to dial in your Midjourney capabilities.

📖 Resources

Here’s a great resource from Chilton Webb on some of the things that MidJourney can do and nice examples, like this:

And a few other helpful articles:

• How to use Midjourney: A complete Guide

• How To Create Stunning Photorealistic Images

That’s it for this week.

What strategies and use cases have you found for your generative AI images?

Stay Super, Marketers!

Gen