How Zapier Transformed Its Marketing Team Into AI-Native Builders

Zapier Chief Marketing and AI Officer Dan Slagen picked a controversial way to start off a team meeting on AI:

He showed Zapier’s marketing team who they'd be competing against for their next job.

Zapier is an automation company that helps businesses connect their apps and automate repetitive work. Zapier's roughly 900 employees work fully remotely, but once a year they gather in person.

When they got together in 2026, Slagen brought in an expert on Cursor, a popular coding agent, who showed the team how he was using AI agents for marketing and even demonstrated how he’d developed his own digital twin.

Then, Slagen dropped a bomb on his team:

"This morning, you didn't even know that this person existed. You didn't know that you are potentially interviewing against this person for your next role at Zapier or outside of Zapier. Now, how do you feel?"

Zapier’s team, by most metrics, was ahead of the curve on AI. They already had broad access to AI tools. Employees were using their tools of choice and building useful workflows around them.

But Slagen felt some tough love was needed.

He had started noticing something in early 2026 that made this progress look less secure. He observed that marketers without traditional development backgrounds had started using tools such as Cursor and Claude Code to build their own apps, dashboards, workflows, and digital twins.

In short, the best marketers weren’t just talking with ChatGPT. They were now evolving into true AI-native builders who were able to turn ideas into working tools, software, and systems.

Slagen believed this represented a profound shift in what it meant to be a marketer. He knew his team couldn’t afford to watch it happen from the sidelines.

So after the meeting, he called a Code Red.

The Code Red Came With a Three-Week Deadline

Slagen didn't end the meeting with a warning and leave the team to figure out the rest.

He paired the urgency with a promise:

"We are going to build with you. We are going to help you every step along the way."

Then he gave everyone three weeks to get up and running on Cursor. Employees who strongly preferred Claude Code, Codex, or another builder could use it, but Cursor became the default starting point for most of the team.

And Slagen meant the whole marketing team. Product marketers. Brand designers. Customer marketers. This wasn't an advanced experiment reserved for a technical subgroup: Everyone had to start building.

Five Coaches Made the Deadline Possible

A mandate this broad can go wrong quickly. People get stuck, hide that they're stuck, and eventually find a reason to return to the work they already know how to do.

Slagen tried to cut that pattern off early. He assigned five internal coaches who were already showing advanced AI skills and asked them to help teammates get set up, troubleshoot problems, and start working.

Then he gave the coaching system an unusual rule: The coaches would not report who contacted them, what anyone asked, or where a teammate was struggling.

That gave employees somewhere safe to ask the questions they didn't want to ask in front of their boss.

Slagen is also candid that the initial Code Red probably worked more through fear than inspiration. He doesn't love that. But the fear got people's attention, while the deadline, common tool, and private coaching gave them a practical way to act.

Over time, Slagen says, the fear started turning into inspiration.

Then Everyone Had to Build Something

Getting Cursor installed was not the finish line.

Once the three-week onboarding period ended, Zapier ran a two-week "Build-On." Team members got together, chose something useful, and applied their new skills to real marketing work.

By the end of it, three tiers of builders had emerged.

The first group was getting started and might have built a dashboard. The second group created a more advanced system or automated loop. The third group built what amounted to a complete software application.

Zapier's events team landed in that third group. It built an app with Cursor and Zapier to manage an upcoming virtual conference expected to involve more than 100 speakers.

That was the moment the Code Red became more than a dramatic meeting or another round of AI training. The team had produced working things that leaders could see, test, and put to use.

Zapier Didn't Expect Everyone to Build at the Same Level

The three tiers could have looked like uneven adoption.

Slagen saw something more useful: a map for how to roll out AI at different levels of maturity.

Zapier could now give each group a challenge that matched its current ability. Beginners could keep building confidence without being measured against people already creating full applications. More advanced builders could take on harder systems instead of waiting for the rest of the team to catch up.

This kept progress from becoming a competition that only a handful of technical employees could win. Each group had a next step.

And the employees who moved fastest weren't always the ones Slagen expected.

"One of our absolute best AI users on the team right now is someone that works in customer marketing," Slagen says. "Not someone that you would necessarily expect, but this person's a real leader on the team. It's amazing what they've been able to build."

Slagen also saw strong work from the brand team. Meanwhile, some employees in more technical roles adopted the tools more slowly because they believed they were already ahead of the curve.

As a result, the Code Red surfaced a new kind of talent map. Job title and technical background didn't reliably predict who would become a strong AI builder. Curiosity, urgency, and a willingness to rethink the work mattered more.

Zapier Lowered the Cost of Learning a New Way to Work

Even motivated employees face a basic problem: The campaign still has to launch. The pipeline target still has to be hit. The brand work still has to get done.

It is easy for leaders to tell a team to learn AI while continuing to demand the same output, on the same schedule, at the same level of polish. Slagen made a different trade.

If someone needed more time because they were producing a deliverable through a new system, he allowed it. If the first version arrived a little rougher than usual, that could be acceptable too.

That permission mattered because Zapier wasn't asking people to complete the same work with one more tool open. It was asking them to rebuild how the work got done.

The Playbook Behind Zapier's Five-Week Shift

Zapier's Code Red worked through a sequence of choices that reinforced one another:

1. Make the skills gap impossible to ignore. Slagen showed the team what advanced AI-native marketing work looked like before asking them to change.
2. Set a short, concrete deadline. Three weeks gave the team enough time to start without giving the old way of working time to reassert itself.
3. Choose a default tool. Cursor created a common starting point while leaving room for employees with a strong reason to choose something else.
4. Provide confidential peer support. Five internal coaches made it safer to admit confusion and get unstuck.
5. Make people build real work. The two-week Build-On converted training into dashboards, systems, loops, and applications.
6. Advance people from where they are. The three tiers gave Zapier a way to challenge beginners and advanced builders differently.
7. Temporarily relax the demand for polish. Employees received room to learn a new production system without pretending the first attempt would be perfect.

The visible result was a marketing team producing things few nontechnical marketers would have attempted a short time earlier. They built dashboards. They created automated systems. One team built an application to manage a complex event.

But the more valuable result was that Zapier could now see who was ready, who needed help, and what challenge each person should take on next.

The meeting showed the team what their future competition might look like. The Code Red made sure they started becoming that competition themselves.

Did you enjoy this transformation story? Go deeper on how AI is reshaping work and business with The Artificial Intelligence Show. Each week, we break down what matters in AI and what leaders should do about it.