Why Brands Are Adding Mess on Purpose

Scroll through any feed for a minute and you’ll notice something odd. Everything looks wonderful. The lighting is flawless, the gradients are smooth, the typography is impeccable, the people are suspiciously symmetrical. And somehow, it all blurs into one.

For most of the last decade, polish was a signal. A slick, minimal identity suggested a business that had invested in itself. It cost money and skill, so it meant something. Generative AI has quietly changed that equation. A flawless image now takes seconds and costs almost nothing, which means flawlessness no longer tells anyone very much.

Designers have noticed. Trend reports for 2026 are full of what they call “imperfect by design”: visible grain, hand-drawn type, scribbled notes, textures that look as if a real person made them with real tools. Alongside it runs a broader push towards authenticity: user-generated content, honest sourcing and visible human judgement.

It would be easy to dismiss this as a passing aesthetic. But the psychology behind it is older and more interesting than any trend, and it carries a warning for businesses that try to fake it.

One Like insight: When perfection becomes free, evidence of a human becomes valuable.

Why handmade feels different

Machines have been able to make things more precisely than people for well over a century. Yet handmade products never disappeared. People still pay more for a hand-thrown mug than a factory one that holds tea just as well. Why?

In 2015, Christoph Fuchs, Martin Schreier and Stijn van Osselaer set out to answer that question in the Journal of Marketing. Across four studies, they found a consistent “handmade effect”: simply describing a product as handmade made it more attractive. The surprising part was the reason. It wasn’t mainly about quality, uniqueness or even effort. People felt that handmade products symbolically carried the maker’s love. Unsurprisingly, the effect was strongest when people were buying gifts for those closest to them.

A related line of research points to something researchers call the effort heuristic. A 2004 study led by Justin Kruger found that people tend to rate things more highly when they believe more time and effort went into making them. As a summary from Monash University’s retail research centre puts it, products that are perceived to take more time and effort are rated as higher quality.

Put those together and the design trend starts to make sense. A visible brushstroke, a slightly uneven line or a handwritten note is a small piece of evidence that a person spent time on this. When almost anything can be generated in seconds, that evidence becomes rarer, and rarer things tend to feel more valuable.

Highlight: People don’t value imperfection for its own sake. They value what it proves: that someone cared enough to make this.

The AI label problem

The flip side of the handmade effect is becoming clearer as AI spreads. In a 2024 study published in the Journal of Hospitality Marketing & Management, Mesut Cicek, Dogan Gursoy and Lu Lu ran experiments with more than 1,000 US adults. Participants saw identical descriptions of products and services, from smart televisions to medical devices, with one difference: some mentioned artificial intelligence and some didn’t.

Across eight categories, the versions mentioning AI were consistently less appealing. As Cicek explained, “When AI is mentioned, it tends to lower emotional trust”, and lower trust meant lower purchase intentions. The effect was strongest for higher-risk purchases such as expensive electronics and financial services, where people already feel uneasy.

This study looked at product descriptions, not brand design, so it doesn’t tell us how customers react to AI-generated logos or imagery. But it fits with the survey evidence we explored in The Difference Between Being Remembered and Being Watched, where trust in AI was falling even as its use was rising. Put simply, many customers currently read “made by AI” as “made with less care”.

That doesn’t mean businesses should hide their use of AI. Concealing it carries its own risks, especially if customers later feel misled. It does mean the visible face of a brand, the part customers use to judge whether anyone cares, is where human judgement matters most.

Highlight: For many customers, “made by AI” currently reads as “made with less care”. That’s a perception you can’t afford in the places people judge you most.

Imperfection you can’t fake

Here’s the irony that most trend reports skip over. Grain overlays, wobbly hand-drawn type and coffee-stain textures can all be generated by AI in seconds. A business that adopts “imperfect by design” as a style, without any real human behind it, is simply making a more fashionable kind of synthetic content. Customers may not spot it immediately. But the whole value of the signal depends on it being true, and signals that turn out to be fake tend to backfire.

So the useful question isn’t “Should our brand look handmade?” It’s “Where in our brand does a real person’s judgement, effort or presence actually show?”

That’s also the most sensible way to think about AI in a brand system. Sydney agency UntilNow, which uses AI tools in its own work, describes them as great for exploring dozens of directions quickly and poor at producing final brand assets that need to be original and strategically right. Its summary is short: “AI is a sketch tool.” Used that way, AI can take on the repetitive, behind-the-scenes work and free people to spend more time on the parts customers notice.

The trend towards user-generated content fits the same logic. A customer’s slightly blurry photo of your product in their kitchen carries something a perfect studio shot can’t: evidence that a real person chose it and was happy enough to share it. As we explored in Who Else Buys This?, the voices people trust most are often the least polished.

Highlight: Imperfection is only valuable when it’s true. A fake wobble is just another kind of polish.

What this means for a small business

This is one trend where small businesses start well ahead. Big brands are spending heavily to look as though a human made something. Small businesses usually just are the humans who made something.

The opportunity is to let that show. Photograph the actual workshop, not a stock image of a generic one. Let the owner’s handwriting appear on a thank-you card. Share the customer’s photo rather than a rendered mock-up. Keep the slightly quirky illustration a friend drew when you started, if customers recognise it. These are not design tricks. They are evidence.

AI still has a useful place. It can help you resize a post, draft a first version of a newsletter or tidy a spreadsheet of customer feedback. The trick is to keep it backstage, and to keep a human hand on anything customers will read, see or hear as “you”.

A few questions can help:

• Where in our brand would a customer see real evidence of a person?
• Are we using “handmade” styling to suggest effort we didn’t actually put in?
• Which tasks could AI take off our plate so we have more time for the parts customers notice?
• If a customer found out how a piece of our content was made, would they feel let down?

Which brings us back to the feed where everything looks perfect and nothing stands out. The answer isn’t to make things worse on purpose. It’s to make sure that, somewhere in what you show the world, there’s something only a person could have made, and that it’s real.

Sources

1. Christoph Fuchs, Martin Schreier and Stijn M. J. van Osselaer, “The Handmade Effect: What’s Love Got to Do with It?”, Journal of Marketing 79(2), 2015 (DOI 10.1509/jm.14.0018).
2. Monash University Australian Consumer and Retail Studies, “Made with love”. Summary of the handmade effect and of Kruger et al., “The effort heuristic”, Journal of Experimental Social Psychology 40(1), 2004.
3. Washington State University, “Using the term ‘artificial intelligence’ in product descriptions reduces purchase intentions”, 30 July 2024. Summary of Cicek, Gursoy and Lu, Journal of Hospitality Marketing & Management, 2024.
4. UntilNow, “6 Brand Design Trends Shaping 2026”, 20 October 2026.