All that really matters in Marketing

Clue: it's not AI

I was listening to the Chief Revenue Office of Rippling, Matt Plank, talk about his belief in actual sales-people, over and above AI SDRs. 

His logic, now contrary to popular Silicon Valley opinion, boiled down to four points, one of which struck me as the most compelling:

“Personalization. This is more important today than three years ago, but it still requires a human to review an AI-generated templated email. For example, an SDR had a template generated by AI. The prospect had posted on LinkedIn that their father had passed, so the AI template said, “Saw your father passed away. Sorry to hear that. You should check out Rippling.” That won’t work.”

At the heart of the issue here is taste. Specifically, a human’s taste for not using death as an entry point for a B2B SAAS commercial process. 

There’s no doubt AI agents can fire off more emails in a week than mere mortals, but that does not make them more valuable. Not even close. 

If you’re in a B2B environment, think of the gold standard stories regarding marketing and sales, like Sam Blond’s “Champagne Delivery” at Brex, that can only come from a human appreciation of what being inundated with cookie-cutter outbound feels like, and how to stand out from the crowd.

If you’re in a B2C environment, think of all the enviable moments of global social cut-through this year: from Charli XCX’s “Brat Summer”, to the “Very Demure, Very Mindful” craze on TikTok, to Taylor Swift’s sell-out tour, to the Stanley cup craze… These are the phenomena every brand would wish for. None of them derive from a volume-focussed Marketing game. They are ideas germinating from rooms of people with hard-won taste.

The kind of taste displayed by Swift’s management (on display in her Miss Americana documentary), by Charli XCX’s creative director, by Jools Lebron, or by the fine folks in Terence Reilly’s team at Stanley: this is taste developed with time, with deep work and hard yards, with reading long essays not instant summaries, with making major mistakes, with winning big, and with meaningful journeys not immediate destinations. 

What you realise by looking at what quantitatively succeeds in the real world — not simply in VC-driven TechCrunch-mania — it always comes back to the people. You also realise that the more that complexity rises, the more platforms and channels that there are, all that happens is the non-fungible people matter more. 

As Scott Belsky, founder of Behance has said:

“Good ideas [are] the exhaust of human experiences and traumas, mistakes of the eye, and uniquely human ingenuity.”

Another concrete example: yes, it’s transformative that you can use a CRM tool to pump out a gazillion emails with personalised avatars in 100s of languages, but who’s putting together the plan? Why do that? Which consumers will it resonate with? The more ubiquitous the tool, the more valuable the tool-master. 

The winners going, as Rory Sutherland has suggested, are people who are willing to slow down in the face of AI, rather than persistently seeking hacks: 

“Now, here’s my point. Most of you, if you were students, wrote essays or something like that as undergraduates, right? Fairly confident to say that nobody’s actually kept them? Nobody re-reads them. In fact, the essays you wrote are totally worthless.

But the value wasn’t in the essay. What’s valuable is the effort you had to put in to produce the essay. Now, what AI essays do is they shortcut from the request to the delivery of the finished good and bypass the very part of the journey which is actually valuable—the time and effort you invest in constructing the essay in the first place.

Similarly, the valuable part of advertising is, to some extent, the process of producing it, not the advertising itself. Because it forces you to ask questions about a business which people mostly never get around to asking: What do we stand for? What’s our function? Who do we appeal to? Who’s our target audience? How do we present ourselves? How do we differentiate ourselves? How do we make ourselves look different and feel valuable to the people who encounter us?”

Put more succinctly by Keith Weightman,”it takes months to become proficient, years to become good and decades to become great.” 

There’s no doubt that the democratisation of knowledge through YouTube, Reddit and the like can compress timelines and open seals to make more people able to learn more quickly, but you can’t cheat experience. Querying and ingesting someone else’s data set will never beat leaning on your own memory structures.

If there’s pure toil then, of course, AI can and should win. For instance, let’s celebrate Walmart CEO Doug McMillon’s approach: "We've used multiple LLMs to accurately create or improve over 850,000,000 pieces of data in the catalog. Without the use of generative AI, this work would have required nearly 100X the current headcount to complete in the same amount of time"

Automate that kind of work, by all means. But if you get AI to do the ideation work, or piece together the big, big idea which land that viral campaign or unlocks a big new B2B opportunity, anything which benefits from earned secrets… If you do that, you will lose out to the competition who know that, all said and done, the team with the tastiest people will win.