Using Generative AI to Facilitate the Key Art Process
PART ONE: ESTABLISHING THE PARAMETER SPACE
One of the things I think about the most when it comes to GenAI is how to use it to improve art production processes.
In the process of running digital creative teams for the past dozen or so years, I’ve seen a strikingly upward, even exponential, trend in the number of assets we need to generate to feed social platforms, streaming platforms, and the like. There’s always less (or the same amount of) budget and more things that you need to do with it, right?
Therefore, I’m always on the lookout for what can be done to apply creative decisions more effectively to dozens, hundreds, or even thousands of assets at once, which can then be distributed as needed according to a marketing promotion schedule or product feature rollout.
In particular, I’ve been really interested in how can we use AI-enhanced software to make the process of creating and versioning the Key Art imagery used in the promotion of TV shows and films more effective, faster — and therefore less costly.
This a complicated problem with a lot of moving parts that can be manipulated; it reminds me of doing differential equations in my college calculus class. So like, it’s kinda fun! ;) Yes, I weirdly love math. I admit it.
Why Key Art?
Streaming companies and television studios spend an enormous amount of time and money on Key Art. The budget for key art can be either high-tens or hundreds of thousands, per title — or even much more, depending on the specific franchise. Think about the fact that streaming companies may add over 500 new titles per year, and well…you start to realize that the representative Key Art imagery for titles is, by itself, easily a multi-million-dollar yearly investment. Key Art is big business. Netflix even personalizes the art its users see to increase engagement.
Why the F* is Key Art So Expensive?
Let me first briefly discuss what Key Art actually is, and give you a little bit of insight on the current processes surrounding it.
Key art serves as the primary visual representation of a TV series or movie — and it’s an essential part of the marketing campaign. It’s used for posters, billboards, streaming imagery, social assets, youTube thumbnails, DVD covers, and more. It’s designed to grab the viewer's attention, brand the content, evoke the right mood, and convey a sense of the story and genre. A casual observer might spend a couple of seconds looking at a movie poster; but dozens of people sat in various rooms for months thinking about the strategy behind what people see and then making that specific image happen.
Here’s a brief overview of the key art process as it stands currently:
Key Art Process
Briefing and Concept Development: It starts with a creative briefing where the marketing team, directors, producers, and key stakeholders outline the project’s core themes, target audience, and desired tone. They may also specify where the art will be used (e.g., digital, print, etc.) and provide references for style (eg: a lookbook). Once the initial briefing occurs, mood boards and concepts are storyboarded out for review.
Photo and Asset Collection: Photoshoots, film stills and screenshots, and/or environmental and background imagery all need to be planned, created, or gathered from existing sources, based on the concepts proposed. Most of the time, a special shoot just for the key art is arranged. Sometimes this shoot includes video for social assets derived from the key art.
Design and Composition: It’s not an exaggeration to say that 100 or more unique art compositions might be created for review during this phase. A unique show logo (Title Treatment) will also be created if it doesn’t already exist, or revised if it does. Taglines will be created and tried out for different strategic purposes. Branding for all associated owners of the content will be strategically applied and revised.
Feedback and Revisions: There are dozens of stakeholders, and therefore the potentiality for an extensive feedback process. Additionally, some higher-visibility titles may warrant focus group or A/B testing on digital platforms to gauge audience reaction.
Finalization and Adaptation: During this phase, final edits are applied and the key art is versioned for different platforms. Depending on the platform this might mean you make a tall, skinny version of the poster; or a wide version, or a square version, or some other non-standard size. It’s not a simple matter of cropping; elements often have the to be rearranged and the canvases professionally extended. There’s so many places and formats that this usually requires multiple inter-departmental teams in addition to multiple agencies to fully finish everything. Not to mention the international version of the art is often very different depending on a number of factors. The heavy, high res source files have to be carefully packaged and delivered to everyone who needs them as quickly as possible.
Rollout and Integration: Finally, the Key Art is launched as part of the marketing campaign. It may appear on streaming platforms, social networks, on-demand services, and in physical formats, depending on the distribution plan. The launch of the key art sets the visual tone for the rest of the marketing materials, including trailers, social media graphics, and website banners. All these pieces work together to build anticipation and maintain brand consistency.
Bet you didn’t expect all that rigamarole was behind that tiny thumbnail on your Disney+ app, huh?
Over subsequent articles, I’m going to explore and dig deeper into how Generative AI can be used in each part of this process to potentially reduce effort and spend. Personally, I think the tendency for GenAI is to focus way too much on the “creative/pretty image” part and not enough on the “painful repetitive process” areas, of which there are many! I want to look at EVERYTHING.
For example, I see great opportunities for AI to help with things like:
Automated image tagging when uploading images to websites and streaming platforms
Smarter image optimization & compression
Canvas expansion
Faster file delivery
I’m also interested in all forms of improvement to the Key Art process, not just AI-driven.
I want to see if it’s possible to envision a Key Art process that would make sense for current times…much of how the Key Art process functions was put in place many, many years ago. I honestly sometimes question if the idea that there’s “one image to rule them all” is even a concept that makes sense in our modern marketing age.
So…to be continued!
If you made it to the end, thanks for reading, let me know your thoughts :)