Will Vibe Coding Kill MarTech Software?

By Stephen Rothwell, June 24, 2026

I made what I am told was the world's first call from an ordinary GSM mobile phone to a satellite phone on the Globalstar network. A few years later I was part of a team using SMS to do things no one had done before — delivering live football scores to people's phones when “mobile data” barely existed. I went on to found a software house pioneering NFC payments, and then Eagle Eye Solutions, whose technology today powers loyalty for some of the world's biggest retailers — running at around 15,000 API calls a second, 24/7/365, across roughly 500 million customers.

I tell you this not to boast, but to make a point. I have spent my entire career being the person who hacks a world-first into existence, fast, before anyone else. So when people tell me that AI and “vibe coding” now let anyone build software at astonishing speed, my answer is: good. They are right. And it still will not kill MarTech software.

Let me be generous, because the hype deserves it. Vibe coding and AI-assisted development are genuinely outstanding. They collapse the time it takes to get an idea out of your head and onto a screen. They are brilliant at wiring best-of-breed products together, at standing up a prototype, at doing the repetitive graft of writing code. If your problem is “I need to see this idea working by Friday,” these tools are a gift.

But here is what I have learned over thirty years: getting someone — or now something — else to cut your code is not new. I have lived through outsourcing, nearshoring, offshoring, the rise of the big consultancies, low-code and no-code. Each arrived promising to make developers redundant. Each was useful. None of them changed the one thing that actually determines whether you end up with a good system: the design.

Building software is like building a house. It does not matter how fast or how cheaply you can lay bricks if there are no foundations and no plans. Whoever lays the bricks — an offshore team, a low-code platform, or an AI — a bad design gives you a bad building. AI is an exceptional bricklayer. It does what it is told, quickly and tirelessly, like a very good graduate engineer. But it does not originate the design. It will not tell you your foundations are wrong. And sometimes the right answer is not the shiny new technology at all — sometimes the old, boring, proven technique is exactly what the job needs. Knowing which is which is experience, not prompting.

I learned this the hard way, early. When we built that first SMS marketing service for Teletext, we had a working prototype. It worked — and it was going to fall over. It was monolithic, and under real load it would have deadlocked on its socket connections. The engineers disagreed with me; the bottlenecks were real. So I made the unpopular call to redesign it: fully multi-threaded, Java frameworks stripped out because they added complexity, raw sockets instead of HTTP because it was an internal system that needed to be fast, every flashy plugin removed. We built something lean and quick. Formula 1 meets a Porsche — not a luxury SUV.

That same instinct is what made Eagle Eye possible. Creating the features was the easy part. Making them run at 15,000 API calls a second, every second, without falling over — that was the real magic. And again and again, I watched major software names claim they could do the same — then fall over at one API limit or another the moment a procurement team tested them under real load. We failed too, at times. The difference was that we had the team and the brainpower to work through the failures and keep hitting new highs. A prompt cannot do that. A graduate with a chatbot cannot do that. Working through failure at scale is the secret sauce, and it is made entirely of experience.

Vibe coding will not kill MarTech software. It kills the demo.

So, will vibe coding kill MarTech software? No. It kills the demo. It makes the part that was never scarce — writing code — almost free, and in doing so it raises the value of the part that was always scarce: the judgment to design systems that are secure, scalable and robust enough to survive contact with millions of real customers.

There is a real risk hiding in all of this, though. When building becomes this cheap, the temptation is to skip the design step entirely — to vibe-code your way to a MarTech stack the same way enterprises spent the last decade bolting one tool onto another. That does not solve the sprawl problem; it accelerates it. Faster monoliths. More impressive demos that will not survive procurement. Tomorrow's technical debt, built today, at speed.

And there is a discipline that sits above even good design — the one most easily lost in the rush to build. It is having a clear vision and strategy for how you want to serve and manage your customers in the years ahead, and making sure every technology decision is aligned to that direction. Decide first what relationship you want with your customers and where you are taking the business; then choose the technology that serves it. The moment it runs the other way — a shiny platform or a fast-built prototype quietly setting the strategy — you have the priorities backwards. Technology should follow the vision, never lead it.

Which is exactly why this matters more now, not less — and it is why my partners and I started Innovate Marketing. Between Uwe, Ory and me, we have the battle scars: the platforms we have built, the programs we have run, the failures we worked through and the highs we hit on the other side. We have no platform to sell you and no implementation to defend. What we have is the experience to ask the questions AI never will. Is this fit for the strategy you are actually committing to? Will it hold at your real-world load? Is the boring, proven architecture the right one here?

That is what our Enterprise MarTech Modernization Review is built to bring: independent, design-led judgment, before you build, rebuild or replatform. Because the tools to build have never been better, and the need for people who know how to design has never been greater. Helping our clients tell the difference is exactly why we are here.