The art of marketing and the art of software have fully fused together

It’s a fascinating time to work in marketing.

It’s also a somewhat dizzying time, with so much change happening around us.

Those were the opening lines of my book, Hacking Marketing, published 10 years ago this month. That “dizzying time” now seems quaint in comparison, right?

My central thesis: as marketing was becoming a thoroughly digital profession — not just digital channels and touchpoints, but everything about how marketing operated was being shaped by software — concepts and practices from the software development world (the OG digital profession!) could be adapted for marketing teams to help them master their new environment.

(The phrase “digital profession” seems so anachronistic now. Digital anything, really. It’s almost like celebrating electricity. “Marketing is an electrified profession! State-of-the-art, man! Flip that light switch again!”)

I think we can declare that caterpillar-to-butterfly metamorphosis complete.

If anything, marketing isn’t just analogous to software development now — it effectively is software development in many of its activities. How marketers are using AI to develop creative, build campaigns, orchestrate engagement, automate workflows, analyze performance, etc., is almost indistinguishable from how software developers are now using agentic coding. The line between Claude Code and Claude Cowork is pretty thin.

A wild example: Treasure Data’s Treasure Code released earlier this week is literally a CLI — that’s command-line interface — for a CDP. You don’t get more programmer-y than working in a CLI. (Hum that to John Mellencamp's R.O.C.K. In The U.S.A.: C.L.I.’ing In The C.D.P. Sorry, I digress.)

About half my book was dedicated to explaining how agile management from software development could be applied in marketing. Of course, marketers, being the creative souls they are, generally resisted the formal structure of agile methodologies such as Scrum. But things that looked a lot like sprint planning, retrospectives, and Kanban boards proliferated across marketing departments.

More importantly, a culture of adaptability took root. Marketing used to follow a rigid “waterfall” style approach of planning and implementation, often with year-long horizons. But the tempo of the market was running at a much faster speed. The malleability software made it technically possible for marketers to adapt to that pace. But they had to change their operating model to harness it.

(Let’s skip the old argument that “agile” and “strategy” are diametrically opposed. Good agile is an operating system for executing strategy in the real world. Good strategy is intentionally informed and updated by those feedback loops.)

But today, marketing’s metabolism has gone from two-week sprints to two-hour reps.

The bottleneck has flipped. In 2016, the fundamental constraint of agile marketing was execution capacity — you always had more ideas in the backlog than your team could build. Prioritization was the critical discipline because doing stuff was expensive.

AI inverts that equation. When execution of many marketing tasks approaches near-zero marginal cost, the scarce resources become strategy, judgment, and taste. The sprint retrospective — whatever you call it — is arguably now the most important part of the cycle. Not “what can we get done this sprint?” but “what did we learn, and what’s actually worth doing next?”

The book introduced the concept of “perpetual beta,” treating marketing touchpoints as continuously evolving works in progress. AI now makes that trivially easy for any single asset. The harder problem is that AI makes it possible for everything to be in perpetual beta simultaneously. Your website, nurture sequences, ad creatives, onboarding flows, social content — all mutating in parallel.

The next frontier in marketing operations isn’t generating iterations. It’s orchestrating coherence across a hundred things evolving at once. One could almost imagine this as a role of a context-as-a-service platform for the marketing domain.

Agile’s real gift to marketing was never its ceremonies. It was its feedback loop. Plan, do, learn, adjust. AI doesn’t change that loop. It just makes it spin so fast the steps blur together. What used to be a sprint is now a continuous conversation — with your tools, your data, your market. The question is whether we’re listening as fast as we’re talking.

By the way, if you’re interested in a modern take on agile marketing, I highly recommend you connect with Andrea Fryrear of AgileSherpas and Melissa Reeve of Hyperadaptive. Brilliant and pragamatic coaches of present-day marketing agility.

Classic marketing is the interplay of media and messages: how and where something appears, what it says. But in a world of digital experiences, there’s a third, nearly infinite dimension: mechanisms. What an experience does, how it behaves.

For experiences, the mechanism is the message. (With apologies to Marshall McLuhan.)

Admittedly, mechanisms aren’t exclusive to digital. A pop-up retail experience or a cleverly designed unboxing has mechanisms too. But digital is where this dimension becomes nearly limitless — constrained only by the cost of bringing ideas to life.

The challenge was that digital mechanisms required software development. And software development meant engineers. Budgets, backlogs, and build cycles kept most marketing teams from ever fully exploring this dimension. Most ideas would have cost too much and taken too long just to try. Most marketers never developed the muscle for thinking in mechanisms in the first place, because so few ideas could be brought to life, iterated on, and learned from. You don’t develop fluency in a language you rarely get to speak.

AI and vibe coding are changing that, fast. When a marketer can describe an interactive experience in plain language and have a working prototype in an afternoon — not a wireframe, not a spec doc, a functioning prototype — the creative space for mechanisms expands dramatically.

An interactive story that adapts to a reader’s choices. A personalized microsite that assembles itself from a prospect’s industry and role. A product tour that responds to what a visitor lingers on. A competitive comparison tool that pulls in a prospect’s actual tech stack. An onboarding experience that reshapes itself based on how a new customer answers three questions.

These aren’t hypothetical. They’re buildable today by marketers who’ve never written a line of code.

This matters because the first two dimensions of marketing creativity — media and messages — are extraordinary on their own. The art of powerful imagery, compelling video, resonant copy, strategic placement — that’s the craft marketers have been honing for decades, and it's never been more sophisticated.

But mechanisms add a dimension on top of all of that. They turn an audience from viewers into participants. And now, for the first time, experimenting with mechanisms is approaching the same cost and speed as experimenting with messages and media.

Marketing creativity is breaking out of flatland into 3D.

Pace layering might be the most useful mental model for understanding what AI is doing to marketing.

Borrowed from architect Frank Duffy and author Stewart Brand, the idea is simple: any complex system is really a stack of layers, each changing at a different speed. Brand applied it to buildings (the site is essentially permanent; the furniture moves weekly) and to civilizations (nature changes glacially; fashion is fleeting). Gartner adapted it to IT architecture. I adapted it to marketing.

The seven marketing pace layers, from slowest to fastest: company, brand, campaign, channel, tactic, iteration, feedback. At the bottom, corporate culture and values evolve over years. At the top, feedback from the market arrives in real time. In between, each layer operates on its own timescale — and the key architectural insight is that the layers need to work together even though they're moving at very different speeds.

Here’s what AI does to this model: it compresses the top layers toward zero.

Tactic, iteration, feedback — the layers that were already the fastest — are now approaching instantaneous. A tactic that took days to produce takes minutes. Iterations that ran on a cycle of days now happen in a single session. Feedback that took hours to collect and interpret can be collected in seconds with synthetic audiences.

But AI is also reaching down into the middle of the stack. Campaign and channel — layers that used to operate on timescales of weeks to months — are getting pulled upward. When AI can generate a full campaign concept, adapt it across channels, and produce all the assets in an afternoon, those layers start moving at speeds they were never designed for.

Meanwhile, the bottom of the stack hasn’t moved at all. Company culture still changes over years. Brand positioning still changes over quarters. These are deeply human layers — shaped by leadership, relationships, trust, organizational identity — and no amount of AI acceleration touches them.

The gap between the top and bottom of the stack is widening dramatically. And that gap is where things will break.

When tactics and campaigns were slow enough, they naturally stayed aligned with brand and company values. The production timeline itself was a forcing function for coherence — there was time to check, review, pressure-test against strategy. Now the upper layers are spinning so fast that alignment can’t happen by default. It has to be designed in.

This is, I think, one of the most important and underappreciated architectural challenges in AI-era marketing. It’s not just “how do we go fast?” It’s “how do we go fast at the top without losing coherence with the bottom?”

The answer might be counterintuitive: the slow layers become more important, not less. When everything above brand is moving at AI speed, your brand positioning, your values, your company identity become the only stable reference points in the system. They’re the constitution that governs an increasingly autonomous operation.

The companies that invest in clear, deeply understood brand foundations aren’t just strategically sound — they’re architecturally sound. They have something solid for all those fast-moving layers to anchor to.

In a world where anyone can generate anything instantly, knowing who you are is the ultimate competitive advantage. It’s the one thing that can’t be vibed.

Still hacking after all these years,

Scott

P.S. If after reading this, you’re thinking of springing for a copy of Hacking Marketing, I certainly wouldn’t turn down the $2 royalty. However, I’d want you to have the right expectations. Reading it today would be a lot like watching You’ve Got Mail — good story (a marketing-meets-technology love story), but set back in a different time.

P.P.S. One more thing. When I had completed Hacking Marketing and was begging soliciting folks to give “jacket quotes” for the back cover, I took a one-in-a-million shot and reached out to Seth Godin. To my utter surprise and delight, he responded, gave me incredibly positive feedback, and provided an amazing quote.

I was floored! Godin is a legend, and he had been an inspiration to me for two decades. But I had never met or talked with him before. “Go make your ruckus,” he said when he gave me his endorsement.

That generosity of spirit from someone so stratospherically successful to a novice writer like myself touched me deeply. I try to pay that forward whenever I can. I also celebrate that spirit whenever I see it in others. And while it’s sadly not universal, it’s out there when you look for it.