I had three messages about this before lunch, all of them some version of “a thousand jobs, should we move on comp”. The answer is no, and the reason is arithmetic rather than judgement. Two numbers were published yesterday and only one of them travelled. Once you put them next to each other, the thing to do about this announcement changes completely.
What Was Actually Announced
On 11 November 2026 Airwallex said it plans to invest more than S$2 billion in Singapore over the next five years and to grow its Singapore workforce beyond 1,000 people. Airwallex plans $1.6B Singapore investment as co-headquarters, targets 1,000 jobs is how TNGlobal headlined it, and Tech Edition carried the same figures in Singapore dollars.
The rest of the detail: an expanded office at Guoco Tower taking the Singapore footprint past 80,000 square feet across four floors once complete, Singapore reaffirmed as one of two global co-headquarters alongside San Francisco, and plans developed with the Singapore Economic Development Board. The company cited 90 per cent year on year revenue growth and a 90 per cent rise in total transaction volume for its Singapore business in the first half of 2026, against a global headcount of more than 2,700 across 27 offices. The named areas for the new roles are product development, engineering and artificial intelligence.
Two things were not published, and both matter more for hiring than anything that was. There is no split of the headcount target by function, so “product, engineering and artificial intelligence” could mean sixty engineers or two hundred, and nobody outside the company can tell you which. There is no phasing either, which means the five-year horizon could be loaded into year one or back-ended into years four and five. Those two unknowns are the difference between an announcement that changes your hiring plan this quarter and one that does not, which is a reasonable argument for treating the whole thing as context rather than as a trigger.
The Number That Did Not Travel
The company employs around 750 people in Singapore today. The target is more than 1,000. That is about 250 net new roles over five years, and at least one outlet reported it exactly that way: Fintech firm Airwallex to invest S$2 billion in Singapore, create around 250 jobs over five years.
Roughly 50 net additions a year, across a whole company rather than only engineering. Gross hiring will be higher than that because people leave and get replaced, but the structural change to the market is one more employer recruiting steadily, not a demand shock. For context, that is a smaller annual engineering intake than a single large retrenchment round removes from the market in a week.
Let me be fair to the announcement: a co-headquarters designation and a five-year capital commitment are genuinely meaningful signals about where senior product decisions get made, and that does affect the seniority mix of roles available here over time. That is a different claim from a thousand openings, and it is the claim worth responding to.
Our expert take
The pattern we see every time a large expansion is announced here is that employers respond to the headline number rather than the net number, and the response is almost always compensation. It is the fastest thing to do and the least targeted. In the accounts we work with, the companies that lose people after an announcement like this are not the ones that paid less. They are the ones whose engineers found out from the news that a more interesting problem existed nearby, and whose manager had not had a career conversation with them in nine months.
The 3 Moves I Would Make This Fortnight
1. Benchmark two roles, not the whole band
The announcement names product, engineering and artificial intelligence. For most employers reading this, that translates to two concrete profiles that overlap: payments and platform engineers who have worked on money movement, and engineers doing applied artificial intelligence work on production systems. Those are the people who could receive a credible approach in the next six months.
So benchmark those two. Not the whole engineering organisation. A company-wide review is the most expensive available response and the least targeted one, because it spends your retention budget evenly across people with wildly different market exposure. The frontend engineer on your internal tools is not being recruited by a co-headquarters expansion, and giving them the same adjustment as your payments lead tells your payments lead nothing about how much you want to keep them. If you need to build the case internally, the structure we use is in our method for an engineering headcount business case in Singapore.
2. Hunt the second-order gap, not the headline employer
Here is the move almost nobody makes. When a well-funded company adds 50 people a year in a market this size, most of those people come from other employers here. Every one of those moves leaves a vacancy at a company with a smaller brand, a slower process and a less compelling story.
That is your pool. Not the candidates choosing between you and a co-headquarters, which is a fight you will mostly lose on brand and compensation, but the engineers who have just watched a strong colleague leave, whose team is suddenly short, and whose manager is now doing two jobs. Those people take calls. They are not in a formal process, they are not fielding four offers, and the thing they want is usually scope rather than money. Approach them in the six weeks after a departure, not in the three months after a pay review.
3. Have the retention conversation before the recruiter does
This is the one that costs nothing and the one teams skip. Take an afternoon, list everyone on your engineering team who matches the two profiles from move one, and for each of them write down what they would say if asked what they are working on next year. If you cannot answer for somebody, that is the conversation to have this week.
The point is not to pre-emptively offer anything. It is that an engineer who has a concrete answer to “what am I building next” is substantially harder to recruit than one who does not, and the cost of finding out is one afternoon. In the departures we have watched follow announcements like this, the trigger was almost never a number. It was somebody realising they had no answer to that question where they were.
Our expert take
The ordering of those three moves is deliberate and it is the opposite of what most companies do. The instinct is to start with money, because it is the lever a hiring manager can pull without asking anyone, and to leave the conversations until somebody has already resigned. In the accounts we work with, the cheapest of the three consistently returns the most: an afternoon of career conversations has prevented departures that a ten per cent adjustment would not have, because the thing being competed for at senior level is usually scope rather than salary. Do the free one first, and you will know which seats actually need the expensive one.
| Move | Cost | Time to act | What it protects against |
|---|---|---|---|
| 1. Benchmark two profiles only | Half a day plus any adjustment | Two weeks | Spending retention budget on people nobody is approaching |
| 2. Work the second-order vacancies | Sourcing time | Next six weeks, ongoing | Competing head on and losing on brand |
| 3. Career conversations with the exposed group | One afternoon | This week | Losing somebody to a better-defined problem, not a bigger salary |
Not sure whether your team is actually exposed
Send us your engineering roster at role level and we will tell you honestly which seats overlap with what was announced, and which ones you can stop worrying about.
Let us discuss itWhat This Does Not Mean
This is where commentary on expansion announcements usually overreaches, so let me mark the limits.
- It is not evidence of a tightening engineering market overall. One employer adding roughly 50 net roles a year does not move a market this size. If your own hiring has got harder, the cause is more likely to be your process or your band than this announcement.
- An announcement is not a set of requisitions. Five-year plans are plans. They get accelerated, delayed and quietly revised, usually without a press release. Nothing here tells you how many roles open in the next two quarters.
- The growth figures are company-reported. The 90 per cent revenue growth and 90 per cent transaction volume increase for the first half of 2026 come from the company, not from an audited filing in the public domain, and the investment figure appears in several currencies and sizes across outlets. Those are reasons to treat the direction as credible and the precision as soft.
- It is not a reason to pivot your product toward payments. Somebody will read a large fintech commitment as a market signal and redirect a roadmap. A single company’s capital allocation is not a demand forecast for your sector. The regional payments picture we wrote about after the Singapore payments network leadership change is a better place to start on that question.
Our expert take
If there is one durable thing in this announcement, it is the co-headquarters framing rather than the money. A capital number tells you about one company’s plans. A decision to make senior product calls here rather than only engineering ones changes what kind of role exists in this market over a five-year horizon, and that is the part worth tracking. It is also the part that is hardest to verify from a press release, so the honest position is interest rather than conclusion.
If your constraint is that you cannot hire 50 people a year against a competitor who can, the structural answer is usually capacity rather than compensation, and our guide to staff augmentation in Singapore covers the trade-offs honestly, including when it is the wrong answer.
Frequently Asked Questions
What did Airwallex actually announce on 11 November 2026?
Does this mean 1,000 engineering jobs are opening?
Should I raise salaries in response?
Where will the realistic hire actually come from?
Two roles to benchmark, not twenty
We help Singapore employers work out which seats are genuinely exposed after an expansion announcement, and shortlist from the pool that will actually take the call.
Let us discuss it