Volume 110: I have seen the future. And it’s...
1. I Have Seen the Future… And It's a Cat In a Shark Suit.
tl:dr: Midjourney, Dall-E 2, and the future.
Last week, my wife and I played around with Midjourney, one of the new AI image generators. It’s pretty simple; you log in to their Discord server and tell it using words what you want to see. The below image was created in about 3 minutes by telling it we wanted to see a cat in a shark suit. Of those 3 minutes, the bulk was processing time.
First, it’s incredible to see something with this degree of fidelity springing from something as simple as a single-sentence command. While I understand there are legal and ethical issues surrounding such creations (who owns the IP, for example) and that some believe it dooms us to a future where nothing is new; we simply recycle what the AIs have learned from the past, there’s no doubt in my mind that tools like this are going to transform the world of design. (And to the originality naysayers, all I have to say is “look around you.” Human originality is usually a mashup of things that have come before, tilted in new and exciting ways. I can’t imagine the use of AI tools changing that in any material way, other than perhaps accelerating it.)
It certainly wouldn’t be the first time tools have changed design. I remember way back when talking with a production specialist who liked to wax lyrical about the days of manual typesetting, which even back in 2000 seemed utterly anachronistic. Since then, of course, we’ve seen the rise of what is now the Adobe Creative Cloud and, more recently, the hyper-growth of tools like Canva, Figma, or even the likes of Squarespace.
And, in much the same way that the Adobe 3D command, introduced in 2003, led to a fad of logo dimensionalization (ironically, the fashion of the moment is to de-dimensionalize), and the gradient tool led to…gradients, new tools, especially AI-based tools, are almost certain to transform what’s possible in brand design.
The question now isn’t whether it happens; it’s when. Let’s take a look at a few likely occurrences (Please take with a pinch of salt, if there is one thing we can predict with a great degree of accuracy, it’s that most predictions turn out to be false)
Craft will decline, ideas will rise.
When speaking with designers about this, I use the statement “you can’t out-kern a machine.” What I mean is that while designers spend years learning, honing, and perfecting their craft, it’s also the first thing AI tools will commoditize as it’s eminently teachable. As a result, we’ll likely see widespread automation of those design tasks that require craft but little creativity. On the flip side, imagination and conceptual creativity will rise to the top. While machines can craft easily and cheaply, they lack imagination. This means the value will shift from craft toward the idea, the creative concept, and the inherent taste of the designer in curating whichever of the machine’s creations are most distinctive, interesting, and appropriate.Commoditized design will be automated out of business.
For years, I’ve failed to understand how some design businesses stay in business with business models utterly dependent on manual labor. For example, look at any design studio focused on CPG/FMCG work, and notably, profit pools such as designing the mechanicals for packaging SKUs bear more than a passing resemblance to Victorian factory work. All of this is likely to be automated in the coming years. It’s increasingly easy to teach an AI different SKU sizes, the graphical elements required, and the basic proportions you want to see them in. From there, all you need is a single human to review the work, make edits, and approve it. If you don’t believe this might happen, Google (and probably others) are already automating UX flows, which can be designed faster and with vastly more test and learn variations than any human.
Directing machines, not just people
Increasingly, the creative directors of the future will be directing machines rather than directing people. Most likely, there will need to be new skills established not just in directing a machine but in directing multiple machines and then curating their outputs into something singular, consistent, and distinctive. And interestingly, in a world where craft is no longer a distinguishing feature of a designer, we may see this new generation coming from wildly different backgrounds and not necessarily with any classical design training.
Now, I know some will say this will never happen. To those people, I suggest they look at what has already happened to lawyers over the past few years, where work is either outsourced to countries with cheaper labor or, increasingly, automated by AI and machine learning systems.
And while there will undoubtedly be pain within this change, I see what’s coming as probably the most exciting period of my career, following what has undoubtedly been one of the least: The fetish of 1950s minimalism in branding design isn’t just dead; it’s a dead end. There’s nowhere left for it to go; it doesn’t differentiate, and it turned an entire generation of brands into a boring melange of nothingness, usually in pastels.
However, the advent of these new tools provides an opportunity to reset for a new future. One where the aperture for ideas and conceptual creativity is pushed wide open and design routes that may previously have been impossible because of the cost of execution, or the perceived complexity of channel fragmentation, or the skill limitations of in-house designers, become eminently possible.
Technology typically does three things–it makes the expensive cheap. It makes the slow fast. And it makes things that were hard easy.
When it does this, those businesses that previously relied on things to be expensive, slow, and hard for their value either adapt or they die.
In the case of brand design, we’ll have to stop looking back with nostalgia at what human beings previously did because the genie is never going back in the bottle. Instead, we need to understand how to harness the machines in such a way as to build new businesses and new value on top. And if history is any guide, these new businesses are often much more exciting than what came before.
And, for those that get there first, they’ll get to transform what we think design can be.
So please don’t screw it up by making everything look like Braun circa 1965. We’ve already done that; it’s boring.
2. Killer Research or Kill the Research Methodology?
tl;dr: Debunking research methods. Finally.
Research is one of those weird fields. Large corporations spend a lot of money on it and then do very little with it, whereas small companies rarely do any.
In the first instance, having a large in-house research group tends to create the Independent Republic of Research, which is vastly more interested in methodological rigor than in solving real business problems–leaving people like me to wade through reams of data and analysis and uninteresting questions in the vain search for something that might be useful. In the second, a whole generation of companies has grown up using behavioral data for insight that views market research as something of a swear word. Why, they reason, would you waste time and money asking people what they think when what people say and do are two entirely different things?
In the past, I’ve referred to this as being “one eye blind.” If all you do is look at research without reference to your behavioral data, you’re blind in one eye. Equally, if all you do is look at behavioral data and do no research, you’re blind in the other. Either way, you’re half-blind.
Anyway, I come to this conversation about research because of a recent article highlighting five research methods that should go away. And I have to say; I found it hard to argue with any of them.
First up, we have consumer co-creation. This had me nodding vigorously and chuckling hard. For years, we’ve listened to branding consultancies dribble on about consumer co-creation as if it’s a transcendent journey into the future of business. But in reality, it’s mostly a crock of shit. First, because consumers rarely care enough to have any interest in co-creating with us, and second, because they rightly wonder why our own experts can’t come up with something better. As Henry Ford once said, “If I asked my customers what they wanted, they’d have told me a faster horse.”
Second, we have “need-state segmentation,” which, if taken literally, is obviously problematic. While it’s beneficial to understand our prospective customer’s needs, the idea that you can then permanently segment people into categories based on these needs is more than a little problematic because the same person often manifests very different needs based on circumstances and context. As an aside, “use cases,” that term so beloved by tech product managers, can at times, be scarily close to need state segmentation.
Next in line is attitudinal segmentation, which is notorious for the people doing it coming up with daft names for the segments. Sigh. (As an aside, I think Apple must’ve hired some of them when it came up with “Dynamic Island”). On balance, I understand where the author is coming from if these attitudinal segments are too small, too niche, and too overlapping. Still, I think there’s a place for it if we avoid the general segmentation trap of boiling things down into too many segments. Unlike the above two methods, I have found value in attitudinal segmentation, even if it’s something you have to read between the lines on, and often find yourself having to jam one or more of them together to get a large enough cohort of people to be interesting.
Fourth, laddering. I have to say I was slightly surprised that laddering still exists because it is, without a doubt, one of the stupidest ideas in marketing (and there’s much stupidity in marketing, folks). To give you a quick overview, it’s an approach where you ladder emotions up to some kind of self-actualization, which might sound OK at face value, but where it invariably nets out is that everything from candy bars to bank accounts to dishwasher soap all make people feel so much like their “best selves” that the brand spontaneously makes people dance, laugh, high-five, and jump in the air in sheer delight. And if that sounds ridiculous, it’s because it is. So, yeah, I’m with the author on this one. Kill laddering. Please.
And finally, we have brand archetypes. Oh my. A favorite pet peeve of mine. Honestly, I’ve always viewed brand archetypes as less a research technique and more branding phrenology. Especially since every strategist/planner likes to make out that the brand they’re working with is one of the cool archetypes, like the magician, the creator, or the explorer, and not the less cool ones like the innocent, everyman, or jester.
I once had someone on a client team adamant that it wasn’t possible to go through a branding exercise without defining the archetype (he was wrong, it’s eminently doable and definitely advisable to go without). Which meant I had to grit my teeth and go through a lengthy process of archetype theater. Never again. It’s rubbish, don’t go there. And if you must, then please ignore the 12 brand archetypes somebody somewhere once made up, and instead make up your own. It’ll be just as useful (e.g., not very), but at least you’ll have a laugh coming up with your own silly names.
3. Too Many Metrics?
tl;dr: Hard to educate the CEO when you’re busy playing defense.
Marketing has many challenges, but undoubtedly one of the most significant is that the C-Suite peers to the CMO, especially the CEO and CFO, typically view marketing as wasteful and a cost to be minimized rather than an engine of growth and profitability to invest in.
There’s also a real irony here that this attitude has been heightened, honed, and then weaponized through the marketing spend of thousands of engineering-led, VC subsidized, mar/ad-tech firms that would have you believe that “traditional” marketing is desperately wasteful and that the answer is the incredible efficiency of “digital” consumer surveillance.
Now, two things have accompanied the rise of this consumer surveillance ecosystem—first, an intense focus on ROI as a metric. Second, an explosion of other metrics that typically have less to do with marketing success and more to do with making whatever ad-tech or mar-tech platform that came up with it look good.
Here’s the challenge with this. First, ROI is a measure of efficiency rather than effectiveness, which means it isn’t necessarily a great choice as the “one metric to rule them all.” Second, the more additional metrics you add and then wallow in, the less likely your CEO or CFO will pay them any attention. More likely, they’ll just default to ROI, creating a Catch 22 for any marketer.
CEOs and CFOs are human beings like any other; they’re busy people, and the idea that they will spend a considerable period figuring out what the 22 metrics on your marketing dashboard actually mean is fanciful, at best.
And yet, we can easily see the perils inherent in placing too much emphasis on ROI. Here are three obvious ones. First, the highest ROI you’ll ever achieve is by spending nothing, which means there’s a huge incentive to underspend in a way that distinctly hampers growth. Tales abound of marketers choosing the less effective path that drew fewer customers because it was more efficient (e.g., higher ROI) than the alternative.
Second, ROI says nothing about your ability to deliver a price premium. The ability to charge more for a stronger brand is a well-known brand effect, yet it’s entirely missing from any calculation of marketing ROI. Incredible when we consider that a 1% lift in price could potentially lift profit by as much as 10%.
Third, focusing only on ROI typically misses that as a business grows, ROI tends to go down as you shift from easier to attract customers to those less inclined to try your brand. Yet, even though this is an essential part of the growth equation, a narrow focus on ROI paints this as a failure rather than a success. This inevitably leads to marketers fishing in a pond of existing customers that are easiest to attract, often offering discounts to try and speed along the transaction, leading to a situation where marketing ROI looks great, but customer growth, profitability, and overall business health are hampered.
So, what to do?
Well, the good news is that there’s a rising tide of empirically researched analysis (assuming we can look beyond the inevitable internecine warfare of clashing egos) that demonstrates how marketing success happens and, as a result, how to measure it.
This suggests that rather than get suckered into the ROI trap or play defense across a broad array of ad-tech and mar-tech vanity metrics, CMOs should instead take the known evidence, apply it to their own business, drop the metrics down to the few that matter, and then go on a major charm offensive with their C-suite peers - showing what should be measured, why it should be measured, and the business results that each metric impacts. Not only that, but they should also demonstrate the risk factors of not doing it this way (e.g., the inevitably harmful business consequences of solely focusing on a metric like ROI)
And, if you find that nobody is listening? That’s fine. You just discovered it’s time to seek a new position months before you’d have gotten there anyway, but with much less pain.