Why taking the AI road could ultimately pay off for OOH creative
By Liseanne Gillham, VP of marketing, Broadsign
Navigating applications of artificial intelligence (AI) in advertising today feels a little like driving on a motorway that's still being constructed. While it may provide a faster route to your destination once complete, the proper exits, overpasses, and signage have yet to be put into place, making it feel like you’re driving a bit blind.
As more brands and agencies are looking for the on-ramp to AI, they’re finding the road isn’t fully paved. For every AI advantage, there are still licensing considerations, privacy and ethics concerns, nascent machine learning models, etc. Generative AI, in particular, is still very early days and it's evolving faster than I can type. But, that doesn’t mean brands, media buyers and owners, agencies, and ad tech developers should shy away from experimentation.
On the contrary, now may just be the best time to dive in and start exploring. If tech companies stopped machine learning (ML) testing when it first surfaced, navigation apps would not be as effective as they are today, facial recognition on the iPhone might not exist, and neither would those content recommendations we’ve come to appreciate across our favorite streaming platforms.
AI has been a part of our daily routines for quite some time, only now, it’s more visible to the mainstream, especially with all the hype around generative AI applications like ChatGPT. Approached responsibly, AI poses a myriad of advantages for the advertising community, and especially for out-of-home (OOH) and omnichannel campaign planning, creative, and execution. It can unlock new possibilities for brands to deliver more targeted, personalized, and unique creative that resonates with consumers across OOH and other ad channels. Look no further than McDonald’s and Burger King’s latest ChatGPT-assisted OOH campaigns as an example.
AI: OOH creative’s friend or foe?
Across the industry, AI experimentation in OOH and omnichannel advertising is already underway, with it serving as a supplemental tool to generate rough concepts for creatives to hone. Even within my own network of marketers, advertising specialists, CG artists, and software developers, I’m seeing wide-spread enthusiasm for generative AI with discussions on how to get the most out of the available tools.
This shift, however, does not come without controversy. To some, generative AI conjures thoughts of a dystopian future in which AI puts us all out of jobs. In reality, it’s highly unlikely computers would ever take over the creative process. As prominent brands like Coca-Cola, Kit Kat, Nike, and Mint Mobile, among others, have experimented with AI-assisted creative and reported pleasantly surprising results, the consensus remains that campaign creative still requires a human touch, and that AI can provide a way to streamline and accelerate non-essential tasks.
How agencies and brands leverage available generative AI tools to advance OOH campaigns will vary, but as a starting point they can help with creative ideation and eventually reduce many of the procedural tasks that consume our time. As teams get started, it’s important to carefully think through your approach, which means doing homework on copyright, licensing, and legal precautions along with exploring available AI toolsets to determine the best fit for each situation. Here are a few of the more obvious areas where we are seeing creative teams experiment with generative AI across advertising that apply to OOH:
Image generation for storyboarding, pitch presentations, and beyond
Compelling visuals are paramount to any OOH ad, and OOH creative featuring dynamic video or animations can make an even bigger splash. Behind every great video or animation is often a well-designed storyboard. Historically, this process has been done manually via hand drawings or digital tools, but generative AI tech is enabling new approaches. Tools like Midjourney, Open AI’s DALL-E, and Stable Diffusion are changing the game, providing an easier way to generate high-quality images for storyboarding and other aspects of the creative process, such as brainstorming sessions and pitch meetings for clients.
In my brief, informal, and totally unscientific LinkedIn poll on generative AI platforms, MidJourney emerged as the clear winner, followed by DALL-E for image generation. One commenter also saw potential in Bing’s New Image Creator. I myself found krock.io’s collaborative AI storyboarding tools to be exceptionally easy to use for rough storyboards.
While these tools are exciting, they also have flaws. A quick internet search on “generative AI hands” will produce hundreds of articles detailing the current limitations of AI art. Still, these tools will improve, and more will follow. As for OOH, AI tools that streamline the creative versioning process will be a big win. It’s possible that in the not-too-distant future, media teams will be able to produce a single OOH creative, with AI analyzing and understanding its various creative elements and then reconstructing it for any OOH screen, regardless of aspect ratio and resolution.
Accelerating copywriting, copy editing, and localization
AI tools like ChatGPT, Jasper, Grammarly, Wordtune, and Copysmith have become more common for enhancing the written word, from copywriting to copyediting and localization. Many companies are taking ChatGPT and these other tools for a creative test run, including our marketing team here at Broadsign.
We have experimented with building social media posts, ad copy, video scripts, storyboards, creative imagery adaptations, and more. The results: it can be challenging to distinguish AI- and human-generated copy, but as a content manager, it’s often hard to trust the results. Fact checking and trusting the accuracy of AI-assisted creative content remain two key struggles as we experiment with AI internally. That said, with time and as AI tools learn and become smarter, I anticipate this may change; AI translation technology can already be used to support text and language translation for global OOH campaigns.
AI warning label: handle with care
AI has certainly become the modern media darling in recent months; a shiny new object that everyone wants a piece of. It’s poised to enhance the creative process, but must be approached delicately. AI tools often produce imagery and copy based on existing materials, introducing potential legal concerns depending on the origin of the source material.
For instance, a handful of Samsung employees accidentally leaked proprietary source code using ChatGPT, and companies like Getty Images are suing AI technology companies for misuse of their imagery to train AI models. Even AI tech leaders like Tesla’s Elon Musk and Apple Co-founder Steve Wozniak called for technology developers to pump the brakes on advanced AI systems for a short duration earlier this year. Furthermore, an AI system’s output is also only as useful as the information it’s fed, so OOH creative teams should prepare for trial-and-error time.
AI may be imperfect and, in its infancy, but it’s poised to transform traditional approaches to OOH creative, and many other aspects of the OOH campaign planning and execution process. There may be no guidebook or map on how to use AI for OOH creative now, but it will soon be a vital tool for the industry. Why not take the road less traveled and start exploring your options?
For the record, this article was written by an actual human, researched using ChatGPT, and edited with Grammarly.