Beyond the Brief: An AI-Powered Ad Campaign

Sifting through stock imagery. Painstakingly matching styles. Manually changing colors and correcting hues.

Unless blessed with original imagery, this was just the way designers had to go about developing a lot of ad creative.

…until AI changed the game.

For D2L Brightspace, an EdTech with bold advertising aspirations, AI was the solution to achieving a consistent yet dynamic visual style all their own, on the double.

As Kenneth Prieto, Creative Team Lead at Superside and the point person for D2L’s campaign, put it, “This approach demonstrates how AI can replace traditional stock imagery, offering a superior creative direction and elevating projects. Not only did we maintain brand consistency, we also injected a fresh, dynamic edge into this campaign.”

So, if like many creatives and marketers you’re on the fence about AI in advertising, let this campaign serve as inspiration. Open your mind, feast your eyes and learn how to make AI your ally.

The Brief: Breaking Free of Drab Ads

D2L wanted to breathe new energy into their brand and bid farewell to the drab, predictable ads of their past.

Inspired by brands like Docebo and The Economist’s iconic advertising where crisp white text slices through bold red backgrounds, D2L envisioned a resonant, eye-catching style.

Their goal: To make heads spin, eyebrows raise and moments people would forever associate with their brand. At the same time, these new ads needed to be made in a simple, efficient manner to enable quick turnarounds on future iterations of the creative.

Taking inspiration from The Economist ads, known for their simplicity and impactful text on a bold color background, the client aimed to replicate this level of resonance and memorability. Their ultimate aspiration was to enter a new era for D2L Brightspace, marked by innovation, playfulness and unquestionably attention-grabbing visuals.

Kenneth PrietoCreative Team Lead at Superside

A tall task, but as it turns out, one that was perfectly suited to a little AI assistance…

Beyond the Brief: Brightening the Future for D2L Brightspace

D2L directed Prieto’s team to “think outside of the box to solve the brief.”

And they did just that—thinking beyond a designer’s traditional toolbox and identifying the opportunity to use AI to achieve the consistent yet dynamic visual style D2L was after. In this instance, Prieto knew stock imagery wouldn’t cut it. Not with D2L’s desire to develop a wholly unique, head-turning style… and not with their timeline to do so.

The client sought a visual style that was both consistent and open to creativity and innovation. Traditional stock imagery, while readily available, often falls short of delivering consistency, uniqueness and freshness. Our challenge was to find a solution that maintained the desired visual style throughout the entire campaign.

Kenneth PrietoCreative Team Lead at Superside

This approach also solved for repeatability. Thinking forward, D2L wanted to be able to quickly turnaround new ads by updating headlines and images. With AI in their corner, Prieto’s team would never run out of appropriate imagery. Once trained on D2L’s style, they could generate new images ad infinitum.

So, how did they go about developing D2L’s desired style to begin with?

They had some direction. D2L provided brilliant copywriting to guide the campaign. They envisioned each ad starting with the tagline “A future this bright…” and featuring minimal, yet magnetic supporting imagery in their orange brand palette. For starters, they shared five compelling headlines:

A future this bright has more degrees than the sun.
A future this bright sends fireworks back to school.
A future this bright nails parallel parks the first time.
A future this bright puts its pants on two legs at a time.
A future this bright turns Eeyore’s frown upside-down.

The brief presented us with five compelling copy lines, all commencing with ‘A future this bright…’ These lines were intended to take center stage in the campaign, and we needed to integrate them with imagery without distracting attention from the main message that was the copy.

Kenneth PrietoCreative Team Lead at Superside

With these concepts in mind, Prieto and his team began experimenting in Midjourney. Their first step was to establish core elements like background color using simple Midjourney prompts. Next, they layered on more words to generate the specific imagery needed to support each headline.

The hardest part? Nailing that first image.

But once they got that first image right, it was a simple matter of keeping the core prompt and changing a few words to swap in different headline concepts. In this way, Prieto and his team were able to achieve consistency while moving fast. He estimates each follow up image took approximately 15 minutes to craft, compared to the hour and a half it would take to source one stock image and edit it in Photoshop.

We experimented in Midjourney with various prompts, meticulously training the bot. Once we achieved the desired style, we made a few prompt adjustments to replicate it across different image concepts. This approach allowed us to maintain fine-grained creative control over the images for each copy line.

Kenneth PrietoCreative Team Lead at Superside

The results? An ad campaign that shines as brightly as the future D2L is helping its customers build. Though it’s too early to share hard numbers, D2L says it’s their best performing campaign in quite some time. While we’re eagerly awaiting the data, let’s look at what we learned.

AI Ad Campaign Takeaways

Lose the fear of AI.

If you take one thing away from this story, Prieto hopes it’s this. He believes AI is here to help, not hurt creative careers. And that, in fact, the best thing you can do for your creative career is learn how to wield it well.

To that end, he shared these words of wisdom:

Just start. It’s easy to start using AI. Anyone can generate an image with a few words. Of course, there’s a learning curve, and if you want to have more control, you need to understand how to use parameters and more complex commands. But at the end of the day, it’s not that complicated and once you start, you’ll learn quickly.
Keep it simple. Single words (or even an emoji!) will produce an image. Keep in mind, short prompts will rely heavily on Midjourney or any AI image generator’s default style, so you need to get descriptive if you want a unique look. That said, super long prompts aren’t always better. Try and concentrate on your main concept.
Forget grammar. The Midjourney bot doesn’t consider grammar, capitalization, sentence structure or words like humans. So, don’t waste time crafting the perfect sentence. Use descriptive words with commas, brackets and hyphens to organize your thoughts, but know the Midjourney bot may not reliably interpret them.
Get specific. Your word choices matter, a lot. The more specific you are, the better. For example, instead of saying “big,” try “gigantic,” “enormous,” or “immense.” Also, use your words sparingly. Fewer words means each one has a more powerful influence, so it’s best to delete filler words whenever possible.
Describe what you want. Not what you don’t want. If you ask for a party with “no cake,” your image may very well include a cake. If you want to make sure an object isn’t in your final image, try using the --no parameter; an advanced prompting technique that allows you to exclude concepts.

D2L didn’t want to tiptoe. So, Prieto proposed they leap forward into the future of design, powered by AI.

If you’d like to join us there, get in touch.