The new AI playbook: The rules brands are setting around customer-facing content
When and how to use AI in advertising imagery is proving to be one of the biggest questions facing marketers this year.
For some brands, AI platforms and services are proving to be game-changing for their ability to speed along creative processes, and save money. Companies like personal care brand Goodwipes are using it not only to create content but also to optimize past photo shoots, while others like Hilton are using AI tools to fuel more ad campaigns without extra shoots. Conglomerates like Unilever are using editing and visualization tools to quickly jump on viral trends, while companies like beverage brand Koia are using it as the basis for entire campaigns that aim to go beyond the bounds of reality.
Canva’s State of Marketing report found that 97% of marketing leaders use AI in their daily creative work, and 99% planned to increase AI investment this year.
But many marketers say they’re careful not to let the tools replace the creative process, and for good reason. Gartner found in June that 49% of U.S. consumers think AI has made content quality worse. That figure rises to 57% among Gen Z and millennials. Brands risk backlash from this cohort if they use AI too obviously, or incorrectly: REI this summer found itself under fire from commenters who roasted the brand for an AI-generated ad of a woman with a bicycle that had two sets of handlebars. Even brands that didn’t use AI in their imagery can get called out for potentially relying on it, as Quip found after a surrealist campaign earlier this year.
The potential pitfalls of using AI mean many brands using it in their customer-facing content are putting guardrails around when and how it shows up.
Mack Reynolds, a former creative strategy lead at Meta, joined the creative services firm Shuttlerock this year as svp of global creative strategy to oversee a team of brand strategists and AI specialists. While AI can be incredibly helpful, he said all brands should consider whether it’s appropriate for the context.
There’s a difference, Reynolds said, between using AI to make templates for product photos that are going to propel a slew of new Meta ads, versus having AI-generated people talk to the camera about why a product is worthwhile.
“Having a brand point of view on your role in people’s lives, and your personality, and your voice, is still the most important thing. And the AI and the tooling you’re using shouldn’t be dictating that back to you,” he said.
Refine, don’t replace, human creativity
One of the most commonly referenced use cases for AI-generated imagery is something that simply is not possible to film in real life. Think: dream sequences, floating products or animated products. Diane DiLorenzo, vp of marketing at plant-based protein beverage brand Koia, said her team has used AI for content that was meant to be larger-than-life or come off as animated, computer-generated imagery. One recent case was its protein powder launch at Costco, which generated 48,000 organic views — about 6.4 times the normal organic views the brand gets on Instagram Reels.
“In those situations, everyone knows it’s AI, and I think that’s another reason why performance is good,” she said. “People are engaging with it and not feeling deceived.”
As a privately owned company that’s getting shelf space in places like Target and Costco as they seek out more protein beverage options, Koia, specifically its marketing team, is using AI to bring ideas to life it may not otherwise have the resources for. In that sense, the tools help smaller brands compete with legacy players, she said.
“AI, in general, isn’t replacing the creative process. It’s just changing it in a big way,” she said. “Early on, we recognized that unique concepts, especially ones on video that are really eye-catching and thumb-stopping for social, could be done through CGI or AI.”
Larger companies are also reaping the rewards of faster processes. Unilever shared last fall that its AI Studio is helping it develop creative assets 30% faster than going without. In one case study, the oral-care brand Closeup produced more than 100 assets across 10 different setups in three days. Unilever also said AI-generated content can outperform its human-made counterparts; it has seen key metrics like video completion and click-through rates double.
Coca-Cola has made similar investments, including drawing headlines last holiday season for using AI in global campaigns. The company’s head of generative AI, Pratik Thakar, said that AI helped the company control every cinematic detail. “This is more than filmmaking — it’s a transformational leap into the future of creativity and technology,” he said in a news release.
‘Remix’ assets IRL shoots to generate fresh content
Reynolds from Shuttlerock said another common use case for AI-generated imagery is beefing up the amount of content brands can put into paid social platforms. That includes tapping back into catalogs of content to refresh them for new audiences, which helps companies lower production costs. He also said it helps monetize the “long tail” of a company, with AI helping pave a lower-cost way to promote lesser-performing SKUs, for instance. “We talk a lot about bringing brand-level craft to your mid- and lower-funnel creative content,” he said.
In one recent case, Shuttlerock worked with Hilton and Meta using AI-powered image-to-video technology. The results were used in travel ads and drove a 42% increase in people reached, yielding an 8% ROAS and 13% more bookings. The company saved the expense of re-shooting at 700 properties. Ultimately, using AI to feed more ad platforms can increase the odds of the algorithm feeding the right ad to the right customer at the right time, Reynolds said.
“It lets us give new life to things that still have plenty of runway, in terms of being able to be relevant and interesting to people,” he said. “You can just update a few elements of it. It doesn’t mean changing anything material about the item, [except] the treatment of it, or the music or voice-over.”
Meredith Diehn, svp of marketing at Goodwipes, said the brand has had success using AI to optimize visuals from existing photo shoots. This can save thousands of dollars, depending on how many days or setups a photo shoot would normally take. She also said click-through rates on AI-created content tend to be above 1%, putting it on the higher end of industry averages.
“We’re a small team, both from a people and resources standpoint. So if AI is a tool that can help with execution, why not lean into it?” she said. “That way, my team can focus more on the strategic level, generating the creative ideas and doing the things AI can’t.”
Set brand-specific boundaries around whether to use AI-created humans
But Diehn from Goodwipes said there are also boundaries for how far AI technology can be used. While the company is comfortable using static AI imagery of people, it’s not using AI-generated actors, for instance.
“We’re talking about a really personal activity; we’re talking about wiping your butt. And there’s a level of trust there. If you have an AI person talking to you about something that’s intimate, do you really trust them? Potentially not,” she said.
DiVincenzo from Koia also said the company hasn’t yet used AI-generated actors in its ads, though she wouldn’t rule it out given the technology’s capabilities and potential future use cases. Still, she said she’s keeping tabs on making sure any AI-generated creative doesn’t read as stale, copycat or generic.
If companies are using the same tools, they risk running into the same type of fonts, aesthetics and even jokes other brands are being fed. Deploying that kind of “sameness” risks diluting the brand’s real message, she said.
“It’s important for the brand to uphold their own brand guidelines and make sure there is a distinct point of view,” she said.
Reynolds from Shuttlerock said whatever makes a brand unique is the message to play up and not water down in the name of efficiency or cost savings. He cautioned against using AI to come up with a whole creative strategy or direction, and instead recommended using it to make the content faster, more relevant and at a higher level than may otherwise be afforded.
“Don’t sacrifice the storytelling,” he said. “Stay true to who you are, but try out a lot of expressions of that and let the creative find the people who are going to tell you what’s most resonant.”
Keep any AI use behind-the-scenes instead of customer-facing content
Despite the upsides of having more content to put out to the world, some brands still choose to keep any AI usage behind the scenes. Jonathan Kantor, chief revenue and marketing officer at high-end pizza oven brand Gozney, said the brand strictly uses AI in drafting processes, and not for anything consumer-facing.
That’s partly because elevated content is a major part of how the company connects with customers, he said. For example, it runs a YouTube channel with nearly 400,000 subscribers that features recipes and outdoor cooking concepts. Customers of the brand, which crossed $100 million in revenue last year, also have a longer consideration journey of about 30 days, with many of its products costing in the four figures. Against that backdrop, the risk of being associated with “AI slop” is too high.
“When you have a highly engaged consumer, and they have an expectation from you, and you’re a premium brand who prides itself on storytelling, AI is a little bit more of a risk right now than an opportunity,” Kantor said.
Kantor said he sees an opportunity for brands who invest in higher-end creative to differentiate themselves as AI content becomes more commonplace. One of the brand’s biggest campaigns this year was a partnership with paint brand Tonester, for which the team built a set so it could douse products in the limited-edition colors. He said some people asked him what kind of AI they used, only to be surprised it was an IRL moment filmed with a camera.
He compared the difference to vintage clothes versus off-the-rack designs, or a chef-driven restaurant compared to a chain.
“I think consumers across the board are going to appreciate the real work more than ever,” he said. “It’s great if brands gain efficiencies and scale to some extent. But if they are artificial, and the quality of work is not there, and it’s not paired with the really, really good, true human stuff, it’s not going to land with the consumer.”