Not if your customer is a living, breathing, emotion-filled person. Our ECD Dom Megna looks at the right and wrong uses of AI in your retail creative.
When it comes to AI in your creative, the temptation to go there is so great. Just the potential of the thing. It’s hypnotising.
As a copywriter and creative director, I’m generally not an insecure creative type. I don’t think AI is coming for my job. Not yet, and if I keep sharp, not ever.
But as a tool for research and discovery? Bring it on.
We’ve been using AI tools across our job functions for a while now. They are kickstarting us on insights, sharpening numbers, and getting us to laborious answers quicker. This is AI doing what it should be doing.
What it shouldn’t be doing is devising or executing a concept to appeal to a human emotion, or spark a visceral response, shock, outrage, or intrigue, or make someone part with their hard-earned money.
As a creative, of course I’m going to shitcan a tool that could do my job. I get it. But it’s pretty clear that the problem with AI as a conceptual tool is that it has no soul. Not yet.
But even more so, apart from a few oldies on Facebook falling for AI Elon images created by his spinbot army, no one is buying it. People look fake, food looks too perfect, headlines are forgettable and copy reads robotically (even when it’s prompted to ‘add personality’).
‘But you can save heaps on those onerous agencies and all that bloated production’, is the common cry.
The rub is that onerous agencies and bloated production are pretty much a thing of the past. Creative agencies and production houses are running as lean as they can before they start to devalue their skills. So basically, all that money you’re saving at one end will leak out the other in uninspired, ineffective comms. It won’t sustain you long term, even though it might look better on this quarter’s P&L.
Even so, there are instances where it can, and should, come in. Laying out a catalogue, re-creating functional product images, creating different angles of those images, writing copy designed just to inform, such as functional eDMs. Those of us in bloated agency land know this and are becoming very good at using it judiciously for these purposes, to the financial benefit of our clients.
Then there’s the whole performance versus brand argument.
Performance is quick and cheap by nature so it doesn’t necessarily need to be good – so AI is fine for that stuff right? Not really. In fact, definitely not.
If you think about your brand being defined by AI, it should make you twitch a little. And the thought of performance just being fodder to test until we get something right and then flog the bejeezus out of it, that just adds to the noise.
Unless you’re so far down the funnel you’re about to land in the sweet, warm world of buying, performance marketing needs a soul as much as brand marketing does. People like something they can relate to – not being treated as mugs. And that performance ad that they’ve been served, it doesn’t marry up to the brand love you’ve been cultivating, then there’s a distinct waft of BS. Your customer doesn’t draw the distinction, they just feel disconnected and that little misstep can cost you someone real fast.
Look, when it comes to AI in the creative department, we’re getting better. Particularly at prompting – this is the human creative touch that AI won’t be able to do without. We’re getting to where we need to be quicker and with more precision, and this will only improve over time.
The ironic thing is that to get any kind of good creative result from AI requires world-class prompting and looking at the wage ranges listed in this Forbes article, a top prompt engineer could cost you up to US$300,000. A lot more than your real-life photographer or director would cost. Naturally, as the tech improves at the same rate of knots it always has, this will correct itself, and I can bet the house your agency is keeping its eye on this.
But in the meantime, if your brand has any kind of soul, don’t sell it for small change. Not yet.
