AI in the hands of a branding guy
AI can give anyone access to branding and marketing tools, but the difference in outcomes relies on who is using them effectively for your brand’s success.
Imagine walking into a shopping centre with thousands of clothes to choose from. You can walk through every store, grab the shirts, pants, jackets and shoes you like, try them on and eventually walk out with an entirely new wardrobe. Or you could walk through those same stores with a professional stylist. The clothes haven't changed. The shops haven't changed. You both have access to exactly the same things. The difference is that the stylist understands what works together, what fits you, what suits the way you want to present yourself, what you already own, what you'll actually wear and perhaps most importantly, what you should probably leave hanging on the rack. That's a round about way I’ve been thinking about how I think businesses should approach AI when it comes to branding and marketing.
AI has given almost everyone access to an extraordinary collection of tools. You can generate an image, write an article, make a video, develop a presentation, brainstorm a name, create an advertisement or produce something resembling a logo within minutes. But access to a tool and expertise in what you're trying to achieve with it are two very different things.
We've already been through this with Canva, Photoshop and countless other tools. Giving someone Photoshop doesn't make them a graphic designer any more than giving me a set of hair clippers makes me a barber. I can technically remove your hair. Whether you want me anywhere near your head is another question. AI hasn't really changed that principle. It's just made the tools considerably easier to use.
AI can make you feel surprisingly good at things
There's an interesting psychological effect that can happen when we make something ourselves. It's been dubbed the IKEA effect, and research into it found that people can place greater value on things they've helped create, even when their creations aren't necessarily as good as they think they are. They conducted experiments involving IKEA products, origami and LEGO, where many participants valued their own amateur creations similarly to expert-made versions simply because they'd invested effort in making them.
From what I can see, AI has become a new version of this. You type the prompt, you choose the image, you ask it to change the headline, you tell it to make the logo bigger, you regenerate it six times until something appears that you really like. And because you were involved in that process, it understandably starts to feel like your creation. I’m guilty of this myself after making personalised birthday cards for my kids. It made me feel, for a fleeting moment, that I was able to illustrate images that look just like them, when all I did was type a couple of sentences.
But there's a distinct difference between that and developing the expertise required to create it yourself, let alone know if it’s objectively good and effective. It's a little like assembling an IKEA cabinet and feeling pretty bloody pleased with the finished result. You did build it, but someone else worked out the dimensions, engineered the joints, chose the materials, designed every component and wrote the instructions that allowed you to put A into B.
AI does that similar kind of invisible work for us. That's not intended to diminish your input either. Knowing what to ask for, making decisions and directing AI can absolutely be valuable skills. But it's worth being aware of the attachment we can develop to something simply because we had a hand in creating it. Especially when someone with experience in that field tells us it isn't working.
The original IKEA-effect research even cautioned that people may value ideas they've invested their own labour into more highly than other people do. So if a designer, marketer, customer or even your own team tells you something looks AI-generated, doesn't feel like your brand or simply isn't very good, don't immediately take that as an insult to your newfound creative abilities. It might actually be useful feedback.
A professional with AI is still a professional
This is the distinction I think gets lost whenever people talk about AI replacing designers, marketers, writers, photographers or brand strategists.
The comparison isn't really: Professional versus AI
It should be a suggestion: You using AI versus a professional using AI
A professional photographer can use computational photography. A designer can use Photoshop's generative tools. A filmmaker can use AI video generation to fill gaps in between what was shot. The advantage isn't that professionals have some secret collection of tools nobody else is allowed to touch anymore. It's that they bring years of experience to the decisions being made before, during and after those tools are used. They know what to ask for. They know what to keep. They know what to change. And perhaps most importantly, they know when not to use it.
That's particularly important with branding because we're not creating pieces of stuff in insolation. A brand needs to work across different environments, over time, for different audiences and across hundreds of touchpoints. The logo you generate today might need to appear on a five-metre sign tomorrow. The character you create might need to appear in 50 different poses. The message you write needs to sound like the same business as the email somebody receives next week.
AI can produce an output, whereas the expertise determines whether that output is actually useful. The distinction I want to make here is that the prerequisite of using AI for branding and marketing isn’t that it requires a branding or marketing professional. But given AI is a tool, and like any tool, when it’s used by individuals that use them every day for specific purposes, it stands to reason that the outcome is going to be more effective when that tool is used by a professional that specialises in the area you need help with and even your specific industry.
It would be no different turning to a professional bricklayer to build a wall, than you going out and buying some bricks and mortar and figuring it out from a YouTube video. One is going to be quicker, more effective and have less headaches, and that’s the price you pay for a professional versus DIY on the cheap, right?
Some things I would not suggest using AI to create
For all the things AI can do remarkably well, there are still parts of your brand where I'd be particularly cautious about asking it to create the finished product.
Logos
A logo is probably the most obvious. Beyond whether an AI-generated logo is strategically or creatively appropriate (which are just as important), there are questions around originality, ownership, copyright and trade mark registration that make generating your primary brand identifier with AI unnecessarily risky. Use AI to research, brainstorm or test directions if it helps, but your final logo should be something you can confidently own, reproduce and protect.
Photos & Videos
I'd also be careful using AI to generate photos or videos of real people, especially those in your team. If you're a dentist, accountant, real estate agent or any other business where customers are going to meet the people they're seeing in your marketing, then credibility matters. Making someone look slightly younger, changing their face, inventing a different pose or generating a synthetic version of them can create that weird feeling that something isn't quite right before a customer has even met them.
Flyers
Then there are flyers and advertisements. AI can certainly help create them, I see them everywhere now. But it's incredibly good at giving you ‘more’. More copy. More graphics. More headlines. More icons. More calls to action. More visual effects. More fluff than the right amount of communication needed to get the point across quickly. The result can technically contain everything you asked for while communicating almost nothing particularly well. If it immediately looks like an AI-generated flyer rather than a communication from your brand, the tool has become more noticeable than the message.
Printing
I'd be especially careful with anything heading to print. Packaging, clothing, stationery, signage and other physical applications require more than generating something that looks okay on your screen. There can be requirements around resolution, colour profiles, bleed, dielines, finishes, separations, artwork setup and the particular production process being used. And if none of that terminology made any sense, there’s a sensible reason for you to consider hiring a professional for this task instead. This is where a graphic designer, finished artist, printer or other production specialist earns their money, because a JPEG isn't necessarily a piece of production-ready artwork.
Some other notable mentions when AI is not the best tool to turn to first are Brand/Product Naming, Packaging Design, and depending on the complexity, any website or mobile app design/development.
Now these aren't arguments for never touching AI. They're arguments for recognising when the stakes of getting something wrong become considerably higher than the cost of getting someone who knows what they're doing to help your brand deliver business success.
Where AI can be incredibly useful
The other side of this is that I use AI myself and think there are plenty of opportunities for businesses to use it well. For instance, AI can be a handy tool for research and data interpretation when it comes to analysing your brand performance. The difference though is that I'd generally use it to help with something in part, rather than it do the whole thing. Along with human oversight of any output AI makes, as it’s easy to miss an extra finger in an image.
So take photography. If you've already invested in professional team photography and someone happens to be wearing the wrong coloured shirt, AI can potentially help alter the clothing quickly. If the background needs extending or changing, it can be brilliant. But I'd try to preserve the actual person. The closer the result stays to the authentic source material, the less chance you have of presenting customers with a synthetic version of somebody they're eventually going to meet.
The same principle applies to brand messaging. AI can be a useful writing partner, but give it something to work from. Have brand guidelines that outline your positioning, key messages, phrases, personality and tone of voice. Feed that context into the process and then actually read the output to make sure your blog articles, social posts, newsletters and documents continue to sound like the same brand. But even still, I’d suggest you take a stab at writing it yourself first and use AI to help you clean up your punctuation, spelling, grammar and succinctness to enhance the effectiveness of what you’re trying to communicate.
Characters and mascots are another good example when needing to re-produce it for low-ROI use cases like a quick social media post, or a last-minute addition to an email newsletter banner image. The point I want to make here is that even though it’s for a minimal impact addition, you still want your mascot to be consistently represented. So rather than asking AI to reinvent your character every time, develop a proper character sheet showing its proportions, front, side and rear views, its expressions, clothing, colours and defining physical traits. The AI then has a reference it can work from instead of guessing what your mascot looks like every time you ask it to wave.
Likewise, if you're generating visuals that contain your existing logo, don't accept some almost-right AI interpretation of it. Your logo already exists. Sub in your actual logo file if needed in a design tool, otherwise make sure it's accurately represented rather than allowing an AI model to decide that today your wordmark apparently needs an extra letter and your icon could do with three additional fingers, if you get what I mean. The takeaway is ensuring the logo is accurately depicted and cross-reference it with your brand guidelines document if you need to.
Another one is an AI-generated website, especially if your business relies on a website or app, like an ecommerce business, this example could be very costly. Now don’t get me wrong, some of them look crazy impressive, especially as website development is not my schtick personally. However, I’ve heard enough tales of people that have made AI-generated websites or apps and over time they used it to fix parts here and there as the site or app has grown. Only to be left with a dog’s breakfast pile of indecipherable code and functionality that an actual website developer needed to be called in to fix to keep the business running. It’d be one thing for a website developer to use an AI tool like Claude Code to develop a website, compared to a gumby like me. As they know what to look for, that might be wrong or better yet, what is easier to use from a consumer user interaction standpoint. While someone like me who hasn’t got the faintest idea if there is a problem with that code, let alone how to fix it, is oblivious when I effectively don’t know what I don’t know when it comes to websites that I don’t work on day in, day out like a developer does.
Lastly, if you're using AI to write your emails for you, especially through systems capable of automatically sending them, put a pause on auto-send and read everything. Every bloody word. And this goes for any form of writing, which could be a social media post or blog article. Your customer doesn't care that an AI wrote something incorrect, inappropriate, generic or weird on your behalf. As far as they're concerned, you sent it. That's ultimately the distinction and that impression you leave someone with might carry a negative impact with it. So AI can assist your brand. It shouldn't be allowed to carelessly impersonate it.
If it looks like AI, that's probably the problem
We're getting better at recognising AI-generated work and our bullshit detectors are on high alert most of the time, especially for those who are very anti-AI. There are certain visual treatments, writing structures, phrases, punctuation habits, overly polished voices, imagery and design conventions that increasingly make people think, “Yeah, that's AI”. And we’ve gotta realise and remember that once that becomes the first thing somebody notices, you've created a branding problem.
Because instead of paying attention to what you're saying, they're paying attention to how you made it. This is why I wouldn't be offended if someone tells you something you've created looks or sounds AI-generated. Even if they're wrong, that's useful information, right? Something about the execution has distracted them enough from the message that the production method became the message.
The same would apply if someone said your brochure looked like a cheap template, your photography looked like stock imagery or your website looked exactly like someone else's. Take the feedback for what it is. Then ask whether the AI has actually improved the communication or simply made producing it easier.
Use AI to help you, not replace you
My general rule for using AI in branding and marketing is pretty simple, use AI to help with something, not necessarily to do the whole thing. Use it to generate options. Extend an image. Help organise your thoughts. Critique an idea. Rewrite something. Hunt for some research and make sense of it for you. Remove a background. Brainstorm headlines. Create variations. Speed up repetitive work. Help turn something that already belongs to your brand into another useful execution. But keep a human making the important decisions. And when those decisions require expertise you don't have, consider whether doing it yourself is actually saving anything or just causing you headaches down the track.
Additionally, to repeat my earlier point, if looks/sounds like AI then it’s going to be an unwanted distraction that detracts from your branding and marketing intentions.
Of course, business owners already have limited time. But you can spend hours prompting AI to develop a flyer, another three trying to make the logo look right, a couple more working out why the printer says your artwork isn't usable and another afternoon rewriting the copy until it sounds vaguely like you.
Hey, you may eventually get there. Or you could spend that time doing the things you're actually an expert at and have someone who understands branding, design or marketing use the exact same AI tools alongside their own expertise to get you somewhere better. That's the opportunity I think AI presents. Not replacing expertise, but amplifying it when used appropriately.
The shopping centre is open to everyone. We can all walk in and buy whatever clothes we like. But if the way you present yourself really matters, there's still considerable value in having someone beside you who knows what suits you, what works together, what you're trying to achieve and what should stay on the rack. AI certainly gives us access to more tools than we've ever had to give your brand a shot at better branding and marketing. Knowing what to do with them to create that better business success is still the important bit.
Comments