Deel grew from zero to $1.5B ARR in seven years while being EBITDA profitable for three consecutive years now. I think all merit is in the outstanding team we have and a few key principles. Some we instilled from the beginning and others we learned through experiments.
I sat down for a few hours to write down my thoughts on how Deel got here and where I see it going. I'll go over:
What most people get wrong about growth and the lightbulb moments that changed how we operate:
- Why selling to large orgs is worth it
- Slack is the office; why founders need to be in the details
- M&A
- Going multiproduct
- Going international from day one
How my job as COO changed as Deel went from 5 to 7,000 people.
AI is supercharging the next growth curve. How we turned Deel into an AI-first business:
- Phase 1: Product-led AI and individual-led AI
- Phase 2: Governance and cross-pollination
- Phase 3: Control token cost
What most people get wrong about growth and the lightbulb moments that changed how we operate
Sell to large orgs even though it's tedious
We started with contractors and SMBs for initial traction. These are self-serve and go live in minutes. That motion is addictive because it's fast.
Enterprise is the opposite. Security reviews, procurement, custom onboarding, and 3-12-month sales cycles.
Every instinct tells you to go where the friction is low. But something being hard and tedious is what makes it less competitive. Especially now with AI, technical moats in software are extremely difficult to build. Doing what's complex in practice builds a moat.
Land only a few large orgs as clients and you get to scale. They add 100s-1,000s of users at once and are very willing to test more of your products if you do a good job. Plus, it's simpler to close one customer than 100.
Slack is the office
We have no headquarters. Slack is where the company actually happens, so everyone at Deel has to treat Slack as they'd do in-person conversation.
If I answer a client in under five minutes, they're still in the chat, so we keep talking. They tell me what's broken and I can fix it. If I reply after a few hours, they'll take the problem elsewhere. The same rule runs internally. When I respond to my team in minutes, they replicate that habit and the whole company moves much faster.
The standard advice is to scale yourself by overseeing groups of people. Though helpful, founders and execs need to be in the details. And I think what people miss about working remotely is that it allows you to be everywhere. Given that everything happens online, you just need to be fast at context-switching and you can influence a huge part of the company.
Alex and Shuo are proof that it can be done. They are able to zoom all the way in and all the way out. All three of us talk with 100+ people on a daily basis, across many initiatives. It's a huge advantage over in-person work.
The first acquisition was a lightbulb moment
Everyone at the company understood what an acquisition was in theory. But knowing the concept and actually buying a business, integrating their people, merging operations, and making it all function, are different things.
Although the first one was scary, it proved M&A was a real, viable route. Fast forward 5 years, Deel has acquired 15 businesses, all of which have contributed massive value.
A lot of the idea behind M&A came from our Chief Strategy Officer, Philippe. He has 30 years of experience and took the company he founded public. Philippe had done M&A before and he pitched it to the rest of us. Time proved him right because he knew the framework to make truly accretive acquisitions. Simply put, growth matters, but profitability makes or breaks a transaction.
Going multiproduct
Thinking your TAM is fixed dooms you. We started with a single product, contractor invoicing. That got us to EOR. EOR to global payroll. Payroll opened HR, then immigration, then the next thing, and the next.
We created most products following customer demand. Once we had thousands of customers and 100k+ users, whenever many of them independently brought up an issue, we knew there was something there.
However, founders need to see the future product and push for better versions. That energizes the whole team, and customers are happy to become contributors. Without Alex and Shuo's vision, Deel wouldn't have expanded as much as it has.
Going international from day one
Most founders treat going global as a sequence. Win your home market, then launch the next country, then the next. The premise is that each market is a separate expansion project. We found that it isn't. We built Deel so any business and any contractor anywhere could sign up from the start. Most platforms only let you sign up from the US, and that caps your business. After knowing there was demand, we added countries to the drop-down list.
Then EOR changed the game because it requires a legal entity in each country. Fortunately, we’d already built a wisdom of crowds early. Just like before, we let clients tell us where they wanted us, and incorporated entities accordingly. This was kind of the easy part, though it can be brutal in places like LatAm.
Operations is the hard part. We made the platform configurable enough to track filing dates, insurance, who runs payroll and how we file with each government. And we ran every new country first through CoreOps, a team of smart generalists whose job was to figure it out, write the SOPs, then hand off to dedicated country teams once it worked.
How my job as COO changed as we went from zero to a 7,000-people, remote operation
I've been COO of Deel since I joined in 2019. Along the way, my role changed dramatically as the company scaled.
The first thing is time horizon expands and shrinks at the same time. As you grow and have more resources, you have to zoom out and make increasingly longer-term decisions. It also shrinks because there are more urgent things to fix, and if you're not in the day-to-day, you start falling behind.
In a 10-person startup, the main job is to execute. At a 7,000-person company, it's to build systems that execute. Initially, you're making nearly all of the decisions, whereas at a large company you're doing internal design, org design, and process design, to push some parts of the decision-making.
Communication is an interesting one. The typical style of exec at a big company is to communicate through leaders. I actually try and talk to as many people as I can, no matter the seniority or tenure. So my communication is very direct, going straight to the IC in most cases.
Hiring profiles change. The people you hire in a small company are very different from the profiles you hire later on. The approach changes, especially for leadership roles. There is more quality and getting to know each other. It takes a lot longer because people begin with more and more leverage. At the start, it's direct and fast.
Metrics also change. I think at a small company you're really focused on getting PMF, whereas at a large company you're focused on hundreds of KPIs underpinned by company goals, and your responsibility is to ensure they're set correctly and people understand them.
Last huge difference: as a small company, HR isn't really a thing. But in a high-growth, medium-sized business, you have to make decisions quickly, which becomes more difficult as the business grows; people will sit on decisions more often. As COO, your job is to preserve startup agility, ensure high performers get promoted quickly, and bottom performers removed.
AI is supercharging Deel's next growth curve
Our approach to AI evolved in three phases.
Phase 1: Product-led AI and individual-led AI.
The product-led approach was straightforward. Engineering teams built AI directly into the product roadmap, creating end-to-end capabilities that ran continuously for customers.
The individual-led approach was different. We encouraged every employee to experiment with AI using ChatGPT, Claude, Gemini, and no-code platforms. These were human-in-the-loop workflows running on people’s laptops and browsers. The goal was simple: create a culture of experimentation and let innovation emerge from the bottom up.
As adoption accelerated, we leaned into both approaches. Every product manager was expected to incorporate AI into their roadmap, and every individual was encouraged to use AI in their day-to-day work. We measured adoption across Deel and actively coached teams that were slower to adopt.
This first phase dramatically increased AI usage across the company.
Phase 2: Governance & cross-pollination
The next realization was that experimentation alone wasn’t enough. To have an impact on Deel as a whole, we needed governance.
We continued giving individuals the freedom to use the tools that worked best for them, but we tracked which use cases were delivering measurable productivity gains and shared best practices across the business.
One thing that was particularly important was celebrating the people doing great work. We gave AI champions a platform to demonstrate what they’d built, share their learnings, and inspire others. That created a flywheel of cross-pollination, and many of those early champions eventually became part of our centralized AI function.
As AI adoption scaled, another challenge emerged: cost.
Phase 3: Control token cost
Token usage was growing rapidly, and we needed a more efficient way to handle it. That’s when we built Akai.
Akai allows us to orchestrate across multiple models, route workloads intelligently, and move many processes from open-ended prompting to deterministic workflows. That has significantly improved reliability and economics.
Throughout this journey, measurement has been critical. We tracked adoption by tool, monitored usage trends over time, and encouraged people to embrace AI. I even found myself sending emails encouraging adoption, which surprised a few people on my team because I almost never send company-wide emails.
Now we’re seeing the results.
Our operations organization hasn’t grown since March of last year, despite supporting significantly more business. Sales teams are selling more, onboarding teams are onboarding more customers, and productivity continues to increase across the company.
We’re still early in the journey, but it’s an incredibly exciting time. AI has moved from experimentation to becoming a core capability across Deel.
I joined Deel because of Alex and Shuo. Everything is execution.
Having ideas is easy. What's rare is finding founders who are able to execute against those ideas for seven years without the pace dropping. Both are still involved in every little detail, and that's truly marvelous to see.
After years of operating at this pace, we have come to expect it, and that expectation is what drives growth. It acts like a standard toward which the team calibrates. Ultimately, principles matter, but only because Deel has an outstanding pair of co-founders.
A piece of advice I'd give is to hire people better than yourself. If you do that, you'll have a very strong team that’s able to tackle any problem. I would also say work to make yourself redundant. Build better processes, products, and hire better people so you don't need to be involved. That's a really good mindset to have for building.
