A long time ago, in an effort to get my advertising students to think about brands in ways they hadn’t (but also, in a sense, had), I asked them to think about what a car would look like if it was created by Apple. Would it look like every other car? Why not? What would be unique about it? What would it do that other cars didn’t? Would it be more like a minivan or a Maserati? Oh and don’t just tell me that it would have a computer in it – that’s missing the point, you fail, get out and go back to math class.
The point of the exercise was to help the students separate their thinking about Apple from the products Apple produced in order to understand what made something uniquely, you know, “Apple”. To try to put some specifics to conversations that invariably were limited to “their stuff looks cool” or “they’re amazing” or “I just really like them.”
There’s an element of that exercise in Brxnd Collabs, a new experiment from Noah Brier. Noah, you may recall, burst onto the public consciousness with BrandTags back in 2008 which, as Fast Company explained, “invited people to ‘tag’ a brand with the first word that came to mind (example: WalMart = big, cheap, etc.), producing a visual ‘tag cloud’ that offered a kind of shorthand, crowdsourced summary of a brand’s meaning.” And even though it sort of started out as a way to test Noah’s friend’s hypothesis that brands live in people’s heads, it ended up being very revealing about just what those brands meant when they were in people’s heads.
“This project” Brier says, “is just many orders of magnitude more.”
In Brxnd Collabs, Brier uses AI to imagine what brand partnerships could look like. Okay, that sounds dull, let me try again. Because it’s not just two brands “shaking hands” and making nice for the camera, which is what most partnerships are. This is more like you put them into the CERN Large Hadron Collider and flung them at each other at nearly the speed of light. They’re more brand “collisions” than “partnerships”.
The interface is simple - you choose two brands from a long list of choices, and also an “item” that you would like to see them expressed on (which, it occurs to me, brings its own sort of “brand” elements to the mix, but let’s ignore that for now lest our heads explode). You click a few buttons and whammo, your collab request goes into the hopper and the AI starts churning away
Because of AI, the results are some sort of hybrid the two brands in question. Sometimes, exciting and innovative combinations of the characteristics of two very different entities that actually make you re-think some of your conceptions about them. And in some cases, the combinations are just 12 car pileups.
In the former camp you’ve got things like the collab between Brooks Brothers and Wrangler. Yes, both clothing brands, but wildly different audiences and even if the audiences do overlap, wildly different ways of engaging with those audiences. Brooks Brothers – founded in lower Manhattan in 1818, has catered to the elite for over two hundred years, outfitting nearly every president since James Madison, and namechecked in works by F. Scott Fitzgerald, Richard Yates and James Thurber.
Wrangler, on the other hand, is a more 20th century brand, one that was overhauled in the 1940s when Philadelphia’s own Bernard “Rodeo Ben” Lichtenstein was hired to design a new pair of jeans specifically for cowboys. He was so innovative and so successful that by 1974, Wrangler was endorsed by the Pro Rodeo Cowboys Association. In the 80s, Wrangler sponsored Dale Earnhardt’s car, and by the 90s, nearly 1 in every 5 pairs of jeans worn was Wrangler.
Both successful. Both authentic. Both very American. Both cater to a distinct audience. But the two could not be less similar on the rack.
And yet the collab… works. It’s a button-down but it’s a casual button-down. Not something you’d wear with a repp tie, to be sure, but also, not something you’d wear if you were busting a bronc. It very cleverly navigates a sort of middle ground between too formal and too casual – while still feeling authentic and real and American.
And all this is very cool and a really interesting (and arguably, billable) way of wasting an afternoon (or three) because you’re not just looking at these partnerships – you’re actually making them yourself. You’re the one deciding which two brands to connect – which means you probably have some personal connection to the brands, they mean something to you so you have a sort of stake in what comes out the other end. Which makes it not only more educational and illuminating, it makes it just more damn fun.
So kudos to Noah and thank you very much. Except…
Except I want more.
Collabs opens the idea that you can take one brand and all the things it means and represents and mix it with another brand and all the things it means and represents, and come up with some potentially, utterly new thing. What I am frustrated by is that the AI platform appears to be looking for patterns that are primarily visual – colors, textures, logos, designs and, um, patterns. It’s not looking at strategy. It’s not looking at concepts.
Take for example, the collab between Lego and Nike. Now of course, I think it’s entertaining that the logo on the front of the shirt says “Liike” – that’s very clever.
And yet, the output is just a t-shirt with a swoosh and some brick-type elements. Come on.
Can we dive a little deeper into these brands to produce something a little more remarkable? Like they’re both about empowerment, right? Whether that’s “Just do it” for Nike or “to inspire and develop the builders of tomorrow” for Lego. But different kinds of empowerment, perhaps. One more physical, one more intellectual (though, yes, I understand that there’s an intellectual component to sport and a physical component to constructing complicated models).
And further, Nike – and of course I’m overstating here, so calm down, Beaverton – is about improving yourself, while Lego is about creating things outside of yourself. And yet because of that, they’re both aspirational. About being vehicles for accomplishing things you thought beyond you.
To be sure, these are all “concepts” as opposed to facts. But Tim Hwang has insisted that AI fundamentally is a concept retrieval system, as opposed to a fact-retrieval system. So this should be right up AI’s street, right?
Indeed, while they are concepts, they are also words in the sentences of a language we use to communicate to each other to help give our lives some meaning.
And that’s what I want. I want to collide the meanings. To find new meanings. Because perhaps when we collide the meanings we will find new insights into the brands people use and thus, in to the very people themselves.
Look, obviously I’m no coder and the ways of AI frighten and confuse me. So what I’m asking for may be ridiculous, absurd, impossible.
But hey, ridiculous, absurd, impossible - that’s AI’s sweet spot, right?