Why are US tech companies building engineering teams in Portugal?

Because they're no longer buying engineering capacity from Portugal, they're building an owned engineering site there. US companies used to buy engineering hours from Portugal on a capacity contract. Now they build an owned second engineering location: a legal entity, a payroll, a P&L line, and a team they intend to keep. Cost still appears in the business case, but it stopped being the reason.

We're Pixelmatters, an AI-native Digital Product Studio in Porto, and we build and run Tech Hubs in Portugal for US companies. So we have a commercial interest in this answer, and this piece argues something that partly works against us: the country is the easy half of the decision, and most of what determines whether it works has nothing to do with Portugal.

What changed between the outsourcing era and now

The old arrangement was a contract for capacity. You described work, a partner staffed it, and the people doing it stayed on that partner's payroll, in that partner's building. If the relationship ended, the knowledge left with them. The location mattered mostly as a line in a rate card.

What US companies are doing in Portugal now is structurally different. They're standing up a second engineering site with its own hiring plan, its own engineering managers, and its own claim on the roadmap. Cloudflare's Lisbon office, announced in July 2019, is the clearest early example: the company evaluated 45 cities across 29 countries before choosing one, and hired for Engineering, Security, Product, Product Strategy and Technology Research rather than support alone. CEO Matthew Prince framed it as access to "a new pool of technical talent that will allow us to bring better products to market."

That's not a sourcing decision. That's a company deciding where a piece of it will live. The failure modes are different too: a capacity contract fails on delivery, an owned location fails on hiring and attrition, and almost nothing written about Portugal addresses the second one.

Which US tech companies already have engineering teams in Portugal?

Enough of them that the precedent is established across a decade and across both major cities. The list below is limited to US-headquartered companies with a stated engineering or product function in Portugal, with a public source for each.

Datadog's hub became operational in July 2022 under Nuno Antunes, VP of Engineering for the Cloud Security Platform, and the company described the plan as growing tenfold within the year. Upwork announced its Lisbon hub at Web Summit on 12 November 2025, its first outside the United States, with CTO Andrew Rabinovich naming AI infrastructure as the reason.

The wider pattern holds. 55 tech companies opened operations in Portugal during 2025, on figures compiled by TeamLyzer: 29 in the Lisbon area, 16 in Porto and the north and center, four elsewhere, six with no fixed site. Treat that count as a directional signal rather than an audited number, since no methodology is published with it.

The table above is deliberately limited to US-headquartered companies. The picture is broader than that, and we've written up which tech companies are choosing Portugal for their hubs across Europe and the US, including Revolut, Zendesk and Oracle's Porto Innovation Center. Worth reading alongside this if what you want is the state of the market rather than the decision behind it.

Why cost, talent and timezone don't settle the decision

Almost every article answering this question gives the same three reasons: Portugal is cheaper than the US, it has good engineers who speak English, and it works US hours. All three are true, and we've argued all three ourselves. Our own piece on why tech companies are opening Tech Hubs in Portugal made the case that companies are now weighing stability and long-term sustainability over short-term cost arbitrage, and we still think that's right.

What that case doesn't do on its own is settle the decision, because cost, talent and timezone are equally true of four or five European countries at once. A shortlist built on them stays four or five countries long.

The numbers behind those claims are real and worth having. Eurostat puts Portuguese hourly labor cost in the information and communication sector at €26.60 in 2024, against Ireland's €70.10 on the same methodology. Portugal ranks 6th of 123 countries on the EF English Proficiency Index. Lisbon and Porto sit within four hours of the US East Coast working day, which only one other country in Western Europe manages at anything close to the cost. We put the full comparison in best countries to build a Tech Hub in Europe, including the criteria where Portugal loses, and the line-by-line spend in what a Tech Hub in Portugal costs.

What the three reasons leave out is everything that happens after month one. They describe a market you're entering. They say nothing about whether you can hire in it, or hold onto the people you hire.

Are Portuguese engineers already working for US companies?

Mostly not yet, and the gap between that and what they want is the reason to move now. The Tech Talent Trends Report 2026, a survey of 1,115 tech professionals in Portugal run between December 2025 and March 2026, found that 80% work for a Portugal-headquartered employer and only 13% currently work cross-border. Foreign employers exist in this market but don't dominate it. Within that foreign slice, the UK is the largest source at 25% and the USA second at 17%.

Now the other number, and it points both ways. 40% say they'd be open to moving to a foreign company, and 19% are actively looking. That isn't a queue of available hires. It's a measure of how loosely people are held in this market, and it applies to the engineers you hire exactly as much as to the ones you hire them from. Local salaries grew 0.9% over the year after seven years of rising, so the market keeps handing people a reason to look, and it keeps handing them that reason after they join you.

That changes what you're building for. The question isn't whether Portugal has engineers who could work for you, and it isn't whether you can beat the local salary, because you can, and it won't be enough. It's whether you get there while the answer to "who do you want to work for" is still open, and whether what you build holds people once every other US company is competing for the same people. An entity in Porto or Lisbon doesn't win that by existing. What it offers that a remote contract can't is colleagues in the same room, a career path with someone local to argue for it, and a team that owns a product rather than a queue. That's a retention instrument. Priced as a cost-savings instrument, it disappoints.

Does opening an office in Portugal attract engineers?

Yes, and this is where the numbers only tell you half of it. A recognized international company opening a real engineering site in Porto or Lisbon is a draw in its own right, and no survey captures it well. Portuguese engineers pay attention to which companies are investing here, what those companies are building, and whether the work is the actual product or the overflow. A named company with a serious roadmap and local ownership of it changes the shape of your candidate pipeline before you post a single role. Cloudflare's John Graham-Cumming put it directly when the Lisbon office opened: "Part of our strength comes from offering a variety of locations to which we can attract people from diverse backgrounds who are excited about working at Cloudflare."

What the data does tell you is what to do with the building once it exists. 62% name "fully remote, with flexible office options" as their preferred work model, another 11% want fully remote with no office, and 2% want to be in an office full time. Hybrid at three or more days a week, the shape most return-to-office policies land on, is the preference of 6%. The report notes that companies are stepping back from flexible policies and pushing fixed office days while candidates still want the balanced version, and it prices the gap: remote-first roles in Portugal pay 14.3% more than office-first roles, €62,798 against €54,951 in average gross salary.

So presence attracts and mandate repels, and they're separate decisions that often get made as one. The investment, the visible commitment to the market, and a team that owns something worth owning are what pull senior engineers toward a new hub. Bolting a five-day attendance requirement onto that spends the advantage you just bought, in a market where 84% of employers already can't fill roles. Scope the office as a place people want to come to and the pipeline compounds. Scope it as a place people have to be and you're competing against a market that pays a premium for the alternative.

How widely do Portuguese engineers use AI tooling?

Widely, and that's the criterion missing from every country comparison in this category. 74% of Portuguese tech professionals use AI coding tools, led by GitHub Copilot at 56%, Claude Code at 34% and ChatGPT Code Interpreter at 31%. Separately, 56.7% report working with AI or machine learning systems in their own products.

This matters more than the hourly rate will in three years, and almost nobody measures it well enough to compare countries on it. What the Portuguese figures say is that an AI-native engineering standard is something you can hire into here rather than something you'd have to install, which is not a claim anyone can currently make about most of Europe with a number behind it.

The report also names the cost of that shift. Senior professionals are becoming substantially more productive with AI assistance, so companies across the market are hiring fewer juniors, which the report raises as a question about how the next generation of Portuguese engineers gets developed at all.

That has a short-term consequence and a long-term one, and they point in opposite directions. In the short term, plan a senior-weighted team: the traditional pyramid, a few leads over a wide junior base trained up across three years, isn't the shape this market is producing right now. In the long term, a market that stops hiring juniors stops replenishing the senior pool that everyone, including you, will be competing for in five years. Both things are true at once, and the second one is the reason we'd argue for keeping a junior intake in a Portugal team even while the arithmetic points the other way. Somebody has to train the people the whole market is planning to hire later.

Hiring speed is the constraint nobody prices

Portugal is a tight labor market, and the cost models circulating for Portugal hubs almost never include the price of hiring slowly. ManpowerGroup's 2025 Talent Shortage survey puts the share of Portuguese employers reporting difficulty finding talent at 84%, against a global average of 74%. Information Technology is the least affected sector surveyed and still sits at 76%.

The supply side is genuinely growing. The European Commission's Digital Decade country report records Portugal's ICT specialists at 5.2% of total employment in 2024, above the EU average of 5.0%, after 13.0% growth in a single year against the EU's 4.2%. The national target is 7% by 2030 and the report calls it achievable. Software developers and analysts account for 51.5% of online ICT job advertisements.

Demand is growing faster. And the pool leaks: roughly 65,000 Portuguese citizens emigrated in 2024, the majority aged 25 to 34, with senior engineering salaries in Germany and the Netherlands running at multiples of the Portuguese band.

So the honest version of the plan is this: you can hire a strong senior engineering team in Portugal, and it will take longer than your model assumes, because you'll be doing it with no local employer brand against companies who have spent years building one. Every month of that is a month the roadmap doesn't move. That cost belongs in the business case, and it's usually the largest number missing from it.

Can your first five hires assess the next twenty?

The first five hires into a new location decide the next fifty, and this is the risk that has nothing to do with Portugal. They set the technical bar, run the interviews, and become the reason the sixth through twentieth hires say yes or don't. Get them wrong and the site doesn't fail loudly. It fills up.

The reason this is hard from six time zones away is that you can't assess senior engineering judgment in a market you don't know, using a network you don't have, against salary bands you're reading off a PDF. US companies that get this right usually do one of two things: they relocate someone who already carries the engineering standard, or they partner with a team that's already hiring in that market and can vouch for the bar.

Cloudflare took the first route. John Graham-Cumming, then CTO and the company's 24th employee, moved to Lisbon to establish the office. That's an expensive answer and a good one. The second route is what a Build-Operate-Transfer engagement is for, and it's the one we run.

What it looks like when it works: UJET's team in Porto

UJET is a US cloud contact center company and a Google Cloud Partner of the Year in 2024. We partnered in March 2025, launched their Portugal operations that September, and the team grew from four people to 25, working as multiple dedicated product teams.

The delivery numbers are the part worth reading. Releases moved from monthly, with a two-week regression cycle, to twice a week, with code reaching production in under a week. Onboarding time improved 4x. The team shipped 16 complex features in the first year, six of them in the first six months, against work that used to wait six to nine months to be scheduled. The end-to-end test suite grew 4.8x, and our team authored 43% of it. All of it is set out in the UJET Tech Hub Case Study.

The reason we can move that fast isn't Porto. It's that the squad comes out of a standing team of 70 or more people who already work together, rather than a recruitment drive with a client's name on it. A traditional hub build starts hiring on day one. Ours starts shipping on day one and hires alongside, which is the difference between a team that's operational in four to eight weeks and one that's operational when the market allows.

When Portugal is the wrong answer

Portugal isn't the right answer for everyone asking this question, and the disqualifiers are specific.

It's wrong if you need to hire 50 senior engineers inside a year. Spain's ICT pool is several times the size of Portugal's, and so are Poland's and the Netherlands'. It's wrong if cost per engineer is the only variable that moves, because Romania and Poland are cheaper. It's wrong for a three-month project, or for one or two developers, where a contractor or Staff Augmentation is the cheaper and better answer. And it's wrong if you never intend to employ anyone locally, because then you're buying capacity, and the whole apparatus of an owned location is overhead you don't need.

Portugal is the right answer for a narrower case than the marketing suggests: a senior product engineering team, working in English on US hours, inside EU employment law, that you intend to still have in three years.

Three routes to a team in Portugal, and what each one costs you

There are three ways a US company ends up with engineers in Portugal, and they differ on who employs the team, how fast it starts, and what you own at the end.

Our own model is the third one. Tech Hub engagements start at €400k a year on a twelve-month minimum, and only the build phase is a commitment: the operate and transfer phases are optional. Transition windows are planned after the first twelve months, and the timing is the client's rather than ours.

Why the country choice is the cheap half

Picking Portugal is the part of this decision that comes with tables, which is why it absorbs most of the attention. It shouldn't. Eight countries reduce to two or three the moment you name the constraint that actually binds, and the tables do that work in an afternoon.

The expensive half is the year after. Whether you can hire senior engineers in a market where nobody has heard of you, while 84% of local employers say they can't. Whether your first five hires can assess the next twenty. Whether the people you hire stay once a US company offers them the same salary to work from home. Whether you'd make the same call in month eight, knowing what you'd know by then.

Spend the diligence there. The country you can decide from a spreadsheet, and Portugal will hold up well in it. What Portugal can't tell you is whether you're building a team or just opening an office, and that's the question the next three years answer.