Tim Armstrong used to milk cows for a living.

Not metaphorical cows. Not "milking the client" cows. Actual, bovine, 5 a.m.-in-the-pouring-rain, moo-ing cows on a dairy farm in rural New Zealand. The kind of place where your nearest neighbor is your aunt who married the guy who grew up across the road, and going to town was a weekly event. Think less Mad Men, more Clarkson's Farm with better scenery and fewer expletives.

Today, he's the Director at Mangrove Digital in Australia, leading digital transformation projects and building audio ad tech infrastructure for major broadcast networks. The career arc from cow paddock to programmatic pipes is, shall we say, not one the LinkedIn algorithm sees coming.

But here's the thing: when I sat down with Tim on the ADOTAT Show, he didn't want to talk about his latest tech stack. He didn't pitch me a platform. He didn't use the word "agentic" even once. Instead, he said something so unfashionable, so painfully analog, that half the ad tech Twitterverse would have spit out their oat milk lattes:

"People are still the cog in the engine."

Reader, I nearly fell off my chair. Not because it's wrong—but because almost nobody in this industry has the guts to say it out loud anymore.

Tim Armstrong's signature analogy is a beauty: "Just because you have a Ferrari in the garage doesn't make you a great race car driver."

He's right, of course. But try telling that to the parade of CEOs who spent 2024 announcing AI-powered everything while quietly laying off the humans who actually understood the business. The ad tech industry has developed a collective delusion that purchasing an AI tool is the same thing as having a strategy. It's like buying a Peloton and calling yourself a cyclist.

Tim's argument isn't anti-AI. Let's be clear about that. He's not some Luddite yelling at clouds. His point is more nuanced and, frankly, more useful: AI is a force multiplier, not a replacement for the thing it's supposed to multiply. You still need human intuition. You still need someone who can read a room, build a relationship, ask the right question—not just a question. AI won't transform a bad question-asker into a good one. It'll just help them ask bad questions faster.

As Tim put it, AI right now can support human thinking, but it can't replace it. Together is the key word. And yet the industry keeps acting like together means "let the robots handle it while I update my LinkedIn bio to say AI Strategist."

Here's a stat that should be taped to the monitor of every CMO who greenlit an AI initiative last year: according to Forrester, 53% of marketers say their company's AI implementation has been underwhelming. Over half. The majority. And that's the people willing to admit it on a survey—the real number is probably higher, because nobody wants to tell the board that the seven-figure AI investment is basically a very expensive autocomplete.

Tim nailed the reason why. It's the same reason it's always been, the one the industry refuses to learn no matter how many times it gets burned: garbage in, garbage out.

"If your infrastructure or the ecosystem within your business isn't ready for AI, then you are significantly inhibiting the opportunity for those initiatives to succeed," he told me. Translation for the C-suite: you can't slap a machine learning model on top of a data infrastructure held together with duct tape and prayer and expect miracles.

Tim sees it constantly in his work at Mangrove. Companies come in desperate to "do AI"—capital D, capital A, capital I—and he has to walk them through the uncomfortable reality that their foundational data, their platforms, their basic operational readiness isn't there yet. That conversation, he says, is a hard one to have, because nobody wants to go back to the executive team and say "actually, before we can do the cool stuff, we need to fix the boring stuff." Boring doesn't get you a keynote at CES. Boring doesn't get you a breathless write-up in TechCrunch. But boring is what actually works.

Here's where things get darker, and where Tim Armstrong's farm-kid common sense becomes genuinely important.

Companies across ad tech—and tech writ large—have been reducing headcount based on AI's promise. Not its performance. Not its proven, measurable, bottom-line-impacting results. Its promise. They're firing people because a pitch deck said they could. And the downstream effect is exactly what you'd expect: the remaining employees are terrified, risk-averse, and absolutely not going to stick their necks out to innovate.

Tim put it bluntly: fewer and fewer people are willing to step out of the mold and say, "I might be putting my job on the line here, but I'm confident we can make this work." Of course they aren't. When your company just laid off 20% of the workforce because a chatbot can allegedly do their jobs, boldness is a career risk, not a career strategy.

This is the part the AI evangelists don't want to talk about. They've created an environment where the very human qualities that make innovation possible—courage, creative risk-taking, the willingness to be wrong—are being systematically punished. You can't automate bravery. And you definitely can't optimize for it in a quarterly earnings call.

One of the most refreshing things about Tim Armstrong is his hiring philosophy. In an industry obsessed with technical credentials and tool proficiency, he cares most about how you think.

He told me about a LinkedIn post from a friend, where people kept reaching out asking "Where do I start?" about getting into a marketing specialty. Tim's reaction was immediate: that's the wrong question. He wants people who show up with a plan. Who've done their research. Who can say, "Here's what I know, here's what I think, here's what I want to do—what am I missing?"

That distinction—between people who ask to be told and people who come with a hypothesis—is everything. It's the difference between an operator and an innovator. Both are necessary, but only one moves the needle. And AI, for all its capabilities, cannot teach someone to be curious. It can't install grit. It can't make someone care about solving a problem they haven't been assigned.

Tim's hierarchy is clear: soft skills over hard skills. Problem-solving over platform knowledge. Cognitive diversity over credential matching. Business acumen can be coached. Technical skills can be developed. But the instinct to look at a broken water pipe on a freezing New Zealand morning and just figure it out? That's either in you or it isn't. And no amount of prompt engineering will put it there.

I asked Tim this question directly on the show, and his answer was devastating in its simplicity: "I think that's what it's become."

Think about that. A guy who works in digital transformation, who builds the infrastructure that enables AI-driven advertising, who has every incentive to hype the technology—and his honest assessment is that AI has become a crutch for leaders who don't want to do the hard work of actually leading.

When ChatGPT removed the barriers to entry overnight, it didn't just democratize access to AI. It created a generation of non-technical decision-makers who got drunk on possibility and lost touch with reality. Everyone became an AI expert. Everyone had a silver bullet. And the people who actually understood the technology—who knew its limits, its requirements, its failure modes—got drowned out by the hype.

Tim's metaphor for this is characteristically direct: AI is like people management. Your team is only as good as the guidance you give them. If your AI strategy is "buy tool, cut headcount, wait for magic," you haven't adopted AI. You've just found a more expensive way to fail.

The uncomfortable truth that Tim Armstrong brings from the dairy farm to the data center is this: there is no technology on earth that replaces the need for a human being who can think clearly, act bravely, and solve problems under pressure. Not ChatGPT. Not your CDP. Not your shiny new agentic AI workflow that your vendor swears will change everything.

People are still the cog in the engine. The sooner this industry remembers that, the sooner we stop wasting billions on Ferraris that nobody knows how to drive.

Coming in Part 2 (Paid): "Audio Is Ad Tech's Dark Horse—And Nobody's Paying Attention." Tim breaks down why podcasting's data problem is actually its biggest opportunity, why Spotify's video push is fragmenting everything, and what the Australian Audio ID initiative means for the global market. Subscribe to ADOTAT to get it first.