AI creative: What gets attention & what builds trust

Successfully leveraging AI-generated creative in your advertising can be tricky.

We’ve taken data points from AI creative presented to real consumers in-feed to show you exactly where and how it works, as well as where it falls flat, especially in terms of building trust

The only constant we found across brands is that there isn’t one. For one brand, even the AI assets that gained the most attention converted at a rate of 0%. For another, AI-generated assets worked all the way through, converting as well as they gleaned CTR and engagement.

Whether or not AI-generated advertising is going to work for your brand is highly dependent on who your consumer is. Keep reading to find out exactly how we tested AI creative across categories and what we uncovered in the wild.

What works about AI creative 

Even with all the advancements in the world of AI, authenticity is still not replicable artificially.

AI technology develops, and even if consumers tend to overestimate their ability to spot AI content (a say-do gap of 66% claiming they felt confident that they could but only 57% actually being able to) skepticism is at an all-time high. According to research from the Nuremberg Institute for Market Decisions, when consumers know either via labeling or their own discernment that ad content is AI-generated, they become generally less engaged and their trust in the brand is negatively impacted.

In a test with a household cleaning brand, we found that creative with an AI-generated hook drove strong engagement, but actual conversion rates were still near-zero, and even the single best-performing asset by way of engagement had zero conversions.

Conversely, for a tech-forward brand, AI creative won decisively on both attention (CTR, engagement) and conversion rate. With such differentiating results, the data here proves that AI creative can work all the way through the funnel when the format fits (in this case, videos for a B2B tool where the pitch is the content).

What doesn’t hold up

Most purchase decisions aren't rational, they're emotional –– the same way a scroll-stopping post beats a well-reasoned ad. 

Based on this data, we’re not able to say definitively that AI usage always drives up engagement, even in individual categories. It won clearly in 1 of 4 tests (tech), was mixed in 1 (household cleaning), and modestly lost in 2 other household brands. Attention went mostly to the human-produced control, not the AI-generated variant.

So, using AI does not guarantee even higher attention rates alone. It’s totally execution and category dependent whether content that’s made with AI will convert or negatively impact your consumer’s trust.

With guidelines around in-feed AI content changing and developing all the time, it’s important to understand exactly how and where to strategize AI usage in your marketing plan. 84% of consumers say they want brands to disclose AI imagery, but when it is, trust drops off. Your target consumer may live in the margins, but launching an AI campaign without confirmation is a serious risk. With labeling requirements meant to build consumer trust causing more scroll-through, knowing your audience and their values is paramount.

AI creative performance is not universal, so brands need to evaluate it by category, execution, and funnel stage rather than assuming AI will automatically drive better results. 

It’s hyper-specific and execution-dependent — our research found that there are no category-wide truths.