- Systematiq AI
- Posts
- Viral Nike x Runway.ml Ad Breakdown
Viral Nike x Runway.ml Ad Breakdown
Welcome to THE party.
I truly appreciate all of you.
You actually make this fun and worthwhile…so
Enough with the sentimental stuff…let’s boogie.
Now, I promised you actionable info…and no bullshit.
So let’s dive right in.
Breaking Perception
NIKE AD LINK ↗️
Last week I posted a “hypothetical ad” for Nike.
The results were polarizing. (That’s what I wanted)
Because we are on the precipice of major change…and you are at the forefront.
And my goal here is to alter your perspective. (Hang with me)
Your Average Marketer:
Your average marketer looks at the ad and says…
“It’s soulless...it has no life”
“I don’t resonate with the message”
“Nike would never air that and you shouldn’t either”
Cool…you totally missed the point.
I’m not saying to take this ad verbatim and run it on Meta…I’m simply highlighting how quickly these tools are moving.
And how a guy with no copywriting, video production, photography, or sound mixing experience can create something like this in less than 10 hours.
BUT how YOU can leverage it.
Taking Action:
Everyone wants Ai to do their job for them…and have a completely finished product by typing 15 words into Discord. (You’re smarter than that)
We’re not there (yet).
BUT
There are other ways to use it.
Hypothetical Situation
Let’s say you’re a brand owner, or you work at a creative agency.
You need to develop a print/static ad for the brand/client similar to what I did for Nike.
Think about how much work goes into conceptualizing an ad like that.
Moodboarding alone takes hours or days.
The goal isn’t to create a finished product…the goal is to maximize efficiency and effort.
Let’s say you have an idea in mind, now you need to translate that idea to a photographer…
Who needs to actually shoot the brand photographs…
Would it be easier and more efficient to…
Give them a mood board of Pinterest posts, stock images, and google photos and say “go take photographs” and make it look like this ambiguous bird's nest of data.
Stylize and conceptualize your vision on Midjourney in 20 minutes…and give them a partially finished product and say…recreate this, in this style.
Option 1, takes days/hours.
Option 2, take minutes.
If that resonates with you… you're in the right place.
Until these tools can create finished products…I want you to think like this.
TIME + ACCURACY = EFFICIENCY.
Actionable Tips
Nike Photography Pic:
Here is the base prompt I used for the Nike ads:
extreme closeup Nike photography, Nike brand advertisement, (# of models) (nationality) (gender) models, dynamic amped pose, streetwear, (color scheme), theme, Leica M10 --ar 16:9 --s 685 --c 35 --style raw
Here’s how I scaled the process for efficiency:
By using curly braces “{}”, the repeat parameter (--r), and the chaos parameter (--c).
Curly Braces “{}”: Give you multiple options for one prompt.
Repeat Parameter (--r): Gives you multiple image generations from one prompt.
Chaos Parameter (--c): Gives you a more varied image grid.
Here’s an example prompts:
Curly Braces Prompt: extreme closeup Nike photography, Nike brand advertisement, {2,3,4} {French, Chilean} {Male, Female} models, dynamic amped pose, streetwear, (color scheme), theme, Leica M10 --ar 16:9 --s 685 --c 35 --style raw
This generates (12) unique prompt generations…each with (4) images.
12 x 4 = 48 images.
Now you created 48 options in one prompt. (Have fun)
Remember, time + efficiency.
Runway.ml Tips:
Use closeup photography: The images render in better resolution. The further away the subject is from the camera, the more distorted it will become.
Image example
Use images that have motion/dynamic angles +lighting: Runway worked best for me using images that had some dynamic poses, angles, and lighting.
Are we partying yet?
Good… I’ll see you next week.
And we’ll chat more about operationalizing tools like Midjourney.
If you want to chat about it…hit me on Linkedin DM’s
Or book some time to chat.
Until next time…
RF