1. Of Airbnb & Franken-Ads.tl;dr: Brands matter. Can’t believe I’m saying this. Again. Sometimes, when you’re looking for something to write about, you get lucky, and more than one article of interest on the same topic happens simultaneously. Fortunately, this week happens to be one of those weeks. First up, we have the CFO of Airbnb on a recent earnings call defending its switch to brand marketing. In case you’re unaware, at the onset of the pandemic a couple of years ago, Airbnb slashed all marketing spending, including hundreds of millions of dollars worth of performance marketing. It then monitored what happened next during this massively unanticipated A/B test, realized that, for Airbnb, there was very little performance in performance marketing, and decided to focus the vast majority of its efforts on brand marketing because this is driving results. We also have a corresponding article focused on the “performance plateau,” backing up the example set by Airbnb, as it points out something I’ve witnessed again and again among firms that focus the majority of their attention on direct response efforts - namely that you reach a point of saturation, where results stall out, and businesses find themselves spending more money just to stand still. What I find interesting relative to the mounting evidence that brand still matters is just how ubiquitous performance marketing has become in the marketing lexicon, how effectively software engineers have reshaped the marketing narrative to promote it, and how compromised this orthodoxy of modern digital marketing is. Let the following sink in for a second, a major player in one of the largest consumer categories on earth needs its CFO to defend a focus on brand marketing on earnings calls. The implication is that the performance marketing propaganda machine has been so successful that investors assume it should be the primary marketing focus for everyone. Here’s a quick summary of why this is wrong. Performance marketing doesn’t mean higher performance; it just means it’s paid for based on a measure of performance, like a click, rather than paying for placement. As a result, it’s become a catchall term for digital direct response advertising, where marketers seek to attribute sales to specific ads to maximize efficiency–via measures like ROI and ROAS. In any market, only 5-10% of customers actively seek to purchase at any given moment. Because it requires direct association with a sale, performance marketing solely focuses on these buyers, but as a result, it suffers from three inherent conceits: That we can find, target, and then persuade this 5-10% of people to buy our product at what might be a fleeting moment in time at what is often the end of their purchase journey. That we will somehow be able to distinguish between those customers who wouldn’t otherwise buy from us from those who would; as Airbnb shows, there’s plenty of evidence that attribution models are wrong. Often, these are sales that would’ve happened anyway, but because someone clicked on an ad along the way, it gets attributed as the reason for the sale. This is why performance marketing has been criticized as a way of giving discounts to customers who would’ve bought from us anyway. That we aren’t accidentally optimizing ourselves for the most price-sensitive buyers, who are often actively looking for the best deal, leaving us dependent on a constant cycle of discounting for business.
Here’s the thing, 93ish% of purchases are made with brands we’ve already heard of. This means that if we aren’t also building awareness and salience of our brands before someone is actively in the market, we reduce our available buyer pool, which leads to growth stalling. Think of it like this. Performance marketing picks low-hanging fruit; brand marketing builds a stepladder to higher branches. Pick all the low-hanging fruit without also building a stepladder, and you’ll eventually run into problems. The only question is one of when not if. Now, this is all becoming so obvious; the real question is, how did we get here? Well, here goes. Marketing, and the subset that is advertising more specifically, has suffered for many years from the perception that it represents money wasted, or at the very least, allocated inefficiently. The Wanamaker dilemma, widely attributed to John Wanamaker a hundred years ago, states that “half my advertising is wasted; the problem is I don’t know which half.” When digital became a thing over the past twenty years or so, engineering-led firms like Google, Facebook, and a gazillion other VC-subsidized ad and mar-tech companies jumped on the Wanamaker Dilemma as a business opportunity: If the world spends $780 billion or so on advertising and half is wasted, this means there’s $390bn worth that can be more efficiently allocated. While the following might sound like an overgeneralization, it isn’t far from the truth: Software engineers have never had much interest in marketing because they don’t understand or respect it. Instead, they want to change it. Why? Because they view themselves as rational actors, uninfluenced by image-making, who buy based on rational, functional factors, and feel entirely driven to purchase the best product to meet their needs, irrespective of who it might be from. So they have, for the past 20 years, been on a mission to transform marketing in this image. It’s also why Google, the world’s largest advertising company, does not self-identify as an advertising company. Instead, advertising is merely the slightly grubby monetization engine that enables Google engineers to do all the cool stuff they do. That’s a significant signal right there. In practical terms, this has led us to a point where what was the tail of the marketing dog - direct response advertising, has become the dog. Ultimately, the reason we are here has nothing to do with solving the Wanamaker dilemma because this approach to advertising is often more wasteful than what preceded it. Engineers didn’t understand or particularly care about how people make purchase decisions. Instead, they were busy pursuing a solution to an engineering problem of quantifying and tracking ads that say, “here’s a discount; buy my cookies.” An approach that had a ready ally in CFOs who already doubted marketing efficacy and received little pushback from marketers who’ve traditionally lacked commercial numeracy and political heft. And let’s face it, no marketer has been fired in the past twenty years for elevating direct response, performance, and programmatic activities above all else, irrespective of the value-destroying potential of taking it too far. As an illustration, here’s an interesting stat. Look at any research on the subject, and it will tell you that creative quality is the number one factor in the effectiveness of any ad. Yet, when we buy online advertising, we’re pushed to spend 99% of our effort, and as much as 50% of our budget, on targeting and only a tiny fraction on the ad itself. Why? Put very simply, targeting is a solvable engineering problem, while creativity is not (Or at least, it hasn’t been up until now. Generative AI has the potential to shift this, so look for engineers getting religion on creative efficacy very soon). So, what next? Who knows how the future will unfold, but there are a few things I can think of. First, we’ll likely see a new understanding, where there’s a broader realization that brand marketing isn’t something that should have to be defended on an earnings call, especially as we shift from a focus on growth to profitability. (Arguably, brand marketing helps maintain pricing power, while performance marketing reduces it). Second, I think we’re arriving at a point of realization that the performance plateau is real. For all its much-vaunted hype, the digital performance landscape can’t get you all the way to your destination. And that while it might give you a jump start in picking low-hanging fruit, eventually, you will need that step ladder to get to the higher branches. Third, we’re going to see a whole generation of digital marketers seek to develop an understanding of brand-building as a means of up-skilling themselves, but because so many of them jumped from engineering to marketing after re-factoring marketing to look like engineering, it’ll be a struggle to get there. And finally, as I alluded to above, look for engineers using AI to get religion on creative quality. But their interpretation isn’t making a whole lot of sense so far. They’re trying to turn something inherently qualitative into something quantitative. Pulling ads apart and testing all the elements separately and rationally before bringing them back together into a Franken-ad, which they’ll assure us will be the highest possible performing ad you can place. Only it won’t be because we’ve been here before. It was called “direct mail,” and it died off when response rates plunged below half a percent. So, far from the best creative in history, it’ll just be a mess, like junk mail still is. Get used to it; a precipitous drop in the cost of production due to generative AI means we’re going to see a lot more Franken-ads in the future.
|
