Volume 159: Look Back, Look Forward.
Look Back, Look Forward.
tl;dr: Yeah, it’s that time of year again.
This is likely the penultimate Off Kilter this year, so it’s inevitably the time to look back at 2023 and forward to 2024.
I’d say the past year has been dominated by the continued impact of high inflation and high-interest rates to combat it. At the beginning of the year, corporations were playing a game of publicity ninjutsu. Publicly apologetic about prices having to rise because of supply chain issues while crowing to investors on earnings calls about their ability to raise output prices higher than input costs, increasing margins in the process.
The impact has been to, at least partly, fuel record corporate profits from 2022 into 2023 alongside persistently high inflation that, while it looks to have stabilized for the moment, could certainly spike again should the economy heat up.
The fact that interest rate increases aren’t factored into inflation calculations has largely hidden the extent of a global cost of living crisis, where the poorest half of society faces the greatest degree of distress when both the cost of goods and the cost of borrowing to buy goods go up while their salaries and ability to pay stay largely the same.
Rising interest rates have also had the additional effect of shifting how capital flows within the economy. The easiest way to think of money at a macro-level is that it flows like water toward sources of return. When interest rates are at zero, money left in the bank earns zero, so it flows toward riskier assets with the potential for future returns instead. Because interest rates were at zero for so long, it didn’t matter that these returns might not appear for, say, ten years because the return from alternatives was zero. This is why so much money was pumped into persistently profitless startups over so many years. Today, interest rates are no longer at zero, which means the risk-free return of “money in the bank” is conservatively around 5%. Compound this over ten years and a risky alternative must now return circa 165% to equal the risk-free return. Add the risk premium (e.g., the expectation that we should earn a greater potential reward for taking on markedly higher risk), and the expected reward must be exponentially greater in practice.
This is why capital flowing into VC funds has slowed to a trickle, while the money being distributed by these funds to startups has slowed to a dribble, and pressure has shifted toward the delivery of profits now rather than profits in the future.
We also saw in 2023 that many of the businesses that were funded on the basis of future returns were never likely to become profitable at all. As the poster child for ZIRP (Zero Interest Rate Policy) excess, WeWork burnt over $22bn in shareholder capital to build a business worth roughly negative $3bn when we take the debt it’s attempting to discharge through bankruptcy into account.
But WeWork is just the tip of the iceberg; a chill wind has blown through the non-AI startup world in 2023, where at least $27bn worth of shareholder capital was incinerated by the failure of 3,200 funded startups. And while I feel for the people impacted, I’m far from surprised. Unfortunately, rather than embrace ZIRP as a low-cost mechanism to decarbonize the economy, vast gobs of capital were instead wasted on never-to-be-profitable corporations built atop the fantasy of being technology companies while exhibiting none of the economics of a technology company. (Investors love technology companies because, historically, they’ve had excellent economics. In most corporations, costs and revenue grow roughly linearly, meaning margins stay about the same as a percentage of total revenue. In technology companies, especially software, costs and revenue tend to diverge over time, with revenues growing faster than costs, leading to high gross margins at scale. These high margins can then be reinvested organically into sales, marketing, and engineering, which serves to maintain market position over time, while the increased value of the corporation overall means it can use cheap shareholder capital to grow inorganically via acquisition, with the potential reward of becoming an unregulated monopoly. This is exactly the path Google, for example, took over the years.)
Relative to the worlds of marketing, branding, and design, I’ve long felt that we should be wary of leaning too hard into the activities of the startup community. Without insight into their financial performance, it was always impossible to understand whether they’d stumbled onto a magical formula for growth or were spending shareholder capital like drunken sailors in the pursuit of growth at all costs. Well, now we know; it was a lot less of the former and vastly more of the latter. This is also why, I believe, we’ve started to move away from the startup brand aesthetic of undifferentiated, boring, minimal modernism in pastels, for it can no longer be viewed as the look of success when the corporations draped in it have been reduced to a valuation measured in pennies. The fact that the aesthetic didn’t work because it lacked distinctiveness simply pales into irrelevance by comparison.
Regarding skillsets, I fear 2023 may have begun to expose an entire generation of ZIRP-era marketers as being singularly ill-equipped to deal with its aftermath. Skilled as they’ve become in delivering growth tactics that depend upon cheap capital and unrealistic unit economics within fast-growing categories, without realizing the period represented a huge and artificially stimulated bubble. (A conversation for a future edition).
So, what’s my summary of 2023? While there’s a bunch of stuff we could talk about, from the economic phenomenon that is Taylor Swift to the Saudis buying up sports wholesale to McDonald’s retro-future Starbucks competitor, I think the true underlying tale is of the impact of high inflation and high-interest rates on people and on corporations. Of a cost of living crisis across vast swathes of the consumer landscape and of the death of the profitless corporation.
So what comes next? Well, as I’ve said before, the only thing human beings are consistently accurate at predicting is that our predictions will almost always be wrong. With that caveat in mind, here are three things I think are imminent, if not underway already. (I’m only going to focus on three here, but I have four in mind. SaaS entering category maturity is the fourth, and it deserves its own dedicated edition).
Anyway, here goes. In 2024:
Auto loans will be the canary in the coal mine of a consumer recession
AdTech & Martech will be radically humbled
AI will become ubiquitous yet pedestrian
Let’s take each in turn:
Auto Loans Will Be The Canary in The Coal Mine of A Consumer Recession.
Economic indicators continue to be distinctly mixed in terms of whether we will or won’t see a recession next year. It’s clear the Fed believes it possible, if not probable, since they just announced the likelihood of interest rate cuts rather than rises next year. If a meaningful consumer recession does get underway, it’s likely to show up first in consumer debt, with auto loans being the bullseye. Why auto loans? Well, the pandemic really messed up the market for vehicles. In terms of new cars, supply chain constraints led to auto-makers leaning heavily into models where profit potential was highest, which meant the most expensive and heavily optioned. A lack of supply bled through into the used market, where prices spiked higher than in living memory. Add higher interest rates than anyone has seen in almost twenty years, and you get the double whammy of non-affordability: highly-priced vehicles combined with highly-priced financing. This has driven three effects. First, expensive new vehicles are piling up on dealer lots, and we’re beginning to see a return to heavy discounting, albeit government incentive-fueled. Second, more buyers are now sitting on expensive, long-duration loans than ever before. (Almost 70% of auto loans are on terms longer than 61 months, while 84 and 96-month terms are now common.) Third, expensive financing combined with an increasing supply of new vehicles means used car prices are finally reverting toward the historical mean. If the economy softens further, people will increasingly try to get out from under their expensive auto loans. First, they’ll try to sell or trade the vehicle in; only they’ll find they can’t because they owe more than the vehicle is now worth. When they realize they can’t sell it or trade it in without having to stump up money they don’t have, what follows next is as obvious as it is predictable. They’ll hand the keys to the finance company and say, “Here you go; it’s your problem now; you deal with it.” The threat of personal bankruptcy being the only option when you have a vehicle you can no longer afford to pay for.
So, yeah. If auto-loan providers start truly howling in pain and there’s a commensurate spike in vehicles entering the auction market at firesale prices, then look out. It’ll be the beginning rather than the end of a down-cycle. And if that doesn’t happen, then we’re probably not going to see a bad recession at all. Not right now, anyway.AdTech & Martech to Be Radically Humbled.
Here’s a scary Venn diagram for you. Over the past ten years, exponential growth has led to there being somewhere between ten and twenty thousand AdTech and MarTech solutions out there. Technology now represents 25%-30% of marketing budgets depending on which report you read, second only to staff costs and above both creative agencies and media spend. Meanwhile, Gartner states that MarTech utilization has dropped from 42% to 33% in just twelve months. And, a recent ANA report says the programmatic supply chain for digital advertising is so convoluted that only about 36 cents of each media dollar makes it to a potential customer.
The above numbers are simply incompatible with a continued status quo.
With marketers being asked to tighten their belts and do more with less, technology is already being touted as the next thing on the CMO chopping block, and it’s going to get ugly.
I’ve worked with several AdTech and MartTech firms over the years, and I can confidently state that the category does not fully comprehend what is coming. Rather than adopt the broader B2B shift from vendor toward business partner focused on consultative selling and the delivery of business outcomes, the AdTech and MartTech worlds remain stuck in a race to see who can shout loudest, are decidedly ambiguous about the line between hype and reality, make things overwhelmingly complicated in order to extract excess profits, and typically sell the sizzle over the steak. Worse, there’s a strong tendency toward personal ego, narcissism, and arrogance.
As a result, and almost certainly without realizing it, they’ve rooted themselves firmly in the camp of vendors rather than business partners. And I’ve yet to meet a client unwilling to jettison a vendor when the opportunity arises.
Now, clearly, the above generalization does not reflect every player in the space. Far from it. I’d just say that while overall, there’s been a broad shift underway in the B2B environment to increase value and stickiness by acting as business partners supporting the delivery of key business outcomes, those focused on selling to marketers remain largely in the camp of selling shiny objects.
And in 2024, that’ll almost certainly begin grinding to a distinct halt.
Want to free up some budget to do more with less? Cut your unproductive MarTech spend. Want to increase your media efficiency and effectiveness? Knock a few steps off the programmatic supply chain to shorten it and get more of each media dollar in front of real human eyeballs.
It’s literally that simple. The marketers in 2024 that are the most effective in driving business results from constrained budgets won’t be magicians. Instead, they’ll almost certainly be those that are most aggressive in re-engineering their tech spend to free up incremental dollars to be reallocated toward better creative, superior customer experiences and harder working media.
So, my second prediction is that there’ll be a widespread culling of the shiny object brigade. The weak will be gobbled up by the strong, and the weakest will go out of business entirely. This time next year, the landscape will be radically humbled relative to where we stand today.AI Becomes Ubiquitous Yet Pedestrian.
I love this quote about self-driving cars. Unfortunately, I can’t remember who said it. It goes something like this:
When you first get into a car that drives itself, it’s frightening; after five minutes, it’s exhilarating. And after fifteen minutes, it’s boring.
This almost perfectly encapsulates how I feel about AI. Initially, it was frightening as we all feared losing our jobs. Then it felt exhilarating as we sought to come to terms with new and magical tools like ChatGPT, and then, after a bit, you realize that you’ve now become a bit bored by it.
The problem, I think, is that while AI can be very smart in some ways, its primary applications (at least in my business) seem overwhelmingly pedestrian. It’s truly magical at some things - it can instantly turn this Off Kilter edition into a compelling 10-tweet thread, for example. But no matter how hard I try to force ChatGPT, Claude, or Bard to give me compelling outputs for competitive or brand strategies, they simply don’t. (Don’t worry, these are experiments; I never play fast and loose with client data) No matter how hard I try to make them do otherwise, what they do is spit out the expected category tropes, the obvious and the uninteresting. As a result, the best use I’ve found in terms of creating outputs is as an obvious filter. Put simply, if a GPT suggests it, then it’s probably something so obvious that it should be avoided.
I’m stealing from something a former colleague (hey, Nick) said on LinkedIn about separating work from labor. The gist is that labor is no more than that, while work requires the use of imagination, creativity, strategic reasoning, and experience. I like this parsing because it gets to the heart of an issue I see with knowledge work, which is that much of it isn’t knowledge work at all. More accurately, we might label it manual labor of the mind.
As a result, while I think there will continue to be much AI hype and a whole slew of dedicated AI tools rather than general-purpose GPTs, their most successful implementations will almost certainly be decidedly pedestrian. Things like analysis at scale, writing and designing for content factories, coding websites, and navigating complex bureaucracies. Meanwhile, the meaningful work that truly requires human expertise will be safe. Yes, the pedestrian elements will be increasingly delivered by AI, but the important stuff will still be driven by people.
And this may not simply be an “only for now” answer. There’s increasing evidence that the scaling of AI intelligence may slow precipitously unless hardware, algorithms, AI models, and synthetic data all improve significantly from where they are today. And while I’d never bet against that happening, I’d also be willing to bet that it’ll take a lot longer than anyone thinks if historical patterns are anything to go by.
So, yeah. My last prediction is that 2024 is the year that AI tools become ubiquitous yet pedestrian. They’ll do much to disrupt the aspects of knowledge work that represent manual labor of the mind and not much at all to disrupt knowledge work that represents true work. That’s not to say people won’t lose their jobs because they undoubtedly will. It’s just that we need to pay particular attention to the nature of the jobs that are lost before hitting the panic button. (And yes, in case you are wondering, 2024 is an important year to ensure that your own job can’t be viewed as modern-day manual labor, and if it can, to try and find a new one that can’t).
So, there you have it. 2023 has been a year defined by the twin challenges of high inflation and high-interest rates. And while both will likely recede in 2024, we’ll continue to feel the effects in the consumer economy (watch the auto loan providers). I predict a bloodbath in the AdTech/MarTech bubble, which it isn’t at all ready for. And, finally, I think that while we’ll become accustomed to ubiquitous AI tools disrupting manual labor of the mind, it’ll play much more of a supporting role in what we might consider truly meaningful knowledge work.