I Built a Cross-Border Recruitment Business in an Afternoon

A recruitment firm in Dubai had everything except the people who needed it. Licenses, employer contracts, the regulatory scaffolding. What it didn't have was any way to reach a Colombian nurse who has no idea the firm exists.

That gap became a business in an afternoon. Here's what actually happened, step by step, so you can judge for yourself.

The Napkin

My partner runs a licensed recruitment operation in the Emirates. Employer relationships, compliance, the boring machinery that makes cross-border hiring legal instead of a scam. Solid business. One blind spot: Spanish-speaking professionals across South America, who represent a huge supply of skilled labor and have never heard his name.

So we talked it through. The opportunity, the audience, the exact shape of the gap. Classic "what if we built a platform that..." territory. The kind of conversation that usually dies in the notes app.

Then I opened CEO.ai and kept going.

The Domain Came First

Before strategy, before a single line of copy, I wanted the name. I worked with an agent that thinks like a domainer - not the lazy "is it available" check, but the useful question: what would a Colombian nurse actually type into Google at 11pm when she's wondering if Dubai is real?

We landed on TrabajoEmiratos.com. "Trabajo" is work. "Emiratos" is Emirates. Someone searching trabajo en emiratos in Spanish is basically typing our URL from memory without knowing it exists.

Then I Asked for the Real Strategy

Not a one-page summary. The actual foundation you'd pay a strategy team for. What came back was more detailed than most decks I've seen from firms that charge six figures.

The Ideal Customer Profiles were the part that stopped me. Age ranges. Countries ranked by push factors and English proficiency rates. Psychographic motivations written as sentences a real person would say - "economic transformation, not just a better job." A fear hierarchy with scams sitting at number one, which, if you know this market at all, is exactly right.

It also mapped trigger events. The moments that flip someone from browsing to acting: a currency devaluation, a friend who already left and posted photos, a contract ending at home. That's the difference between traffic and intent, and it was already in the document.

• Countries ranked by real relocation pressure, not guesswork
• A fear hierarchy that put scam anxiety first, where it belongs
• Trigger events that turn a browser into a candidate
• Language and messaging tuned to how the audience already talks

Then it set a North Star Metric: Verified Placements. Not traffic. Not signups. The count of people who actually relocate and start working. Because that's the one number you can't inflate with ad spend or a good landing page.

"We guide Spanish-speaking Latin Americans through every step of finding, winning, and relocating to a real job in the UAE - for free where we can, affordably where it counts, and safely at every stage."

— The mission statement CEO.ai wrote. I'd put it on a wall.

Then I Built the Site

In Spanish. Shaped around the audience, the brand, and the strategy that had just been written minutes earlier. Not a generic template with the words swapped. The site knew the fear hierarchy, so it addressed the scam question up front. It knew the North Star, so it pushed toward real applications instead of vanity email captures.

Four moves, roughly one hour:

1. 1
   Named it. TrabajoEmiratos.com, chosen for how people search, not how it sounds in a pitch.
2. 2
   Built the strategy. ICPs, North Star, mission. The backbone every later decision hangs on.
3. 3
   Built the site in Spanish. Copy and structure driven by the strategy, addressing the scam fear directly.
4. 4
   Deployed and shared it with people in the target regions. The feedback came back good, and specific.

That last part is what I want you to sit with. This wasn't a mockup I showed a designer. It was live, in the right language, in front of the exact people it was built for, on the same day the idea existed only as a scribble.

Key Takeaways

• Start with the domain if the domain is a distribution channel. For search-driven Spanish audiences, it was.
• Get the strategy on paper before you build, so every page inherits it instead of guessing.
• Pick a metric you can't inflate. It disciplines every decision after it.
• Ship to real users the same day. Feedback beats another planning session.

Next post: how the application layer got built, and how the strategy fed straight into execution there without me re-explaining anything. That part is wilder.

Greg Marlin

Founder, CEO.ai

Greg builds businesses in public to show what the tooling can actually do. When he's not shipping something on a Saturday, he's arguing that most companies are one clear strategy document away from moving ten times faster.