A borderless talent strategy gives you three things a domestic search cannot: access to skills that have gone scarce in every market at once, an honest read on what those skills actually cost, and people who can open a market you have not entered yet. At Olofsson & Company we run these searches from Singapore across APAC and beyond, and the pattern we see is consistent. The firms that get value from borderless hiring are not the ones casting the widest net. They are the ones who map a small, precise population first and treat immigration rules as a design constraint rather than a paperwork step at the end.
Every firm claims a global network. Ask what it actually reaches.
Read the capability decks of the large search firms and you will find the same sentence in all of them: an international network, offices in forty countries, partners on every continent. Fine. Now ask the harder question. When your mandate is a Head of Applied ML who has shipped inference at scale, who does that network actually surface?
Relationship networks reach people who are already inside them. That is their design. A partner calls the people they placed, the people who interviewed well two years ago, the people who came recommended by the people they placed. It is a real asset and it compounds over a career. It also has a shape, and that shape is a social graph, not a labour market.
Here is the uncomfortable part. For a role that has existed for six years in maybe four hundred companies worldwide, the social graph and the labour market barely overlap. The person you want has probably never taken a headhunter's call, because until recently nobody was calling. The network reaches the visible market. Borderless sourcing, done properly, reaches the population.
What a borderless strategy actually buys you
Four things, and it is worth being precise about each.
Access to skills that are scarce everywhere at once. ManpowerGroup's 2026 Talent Shortage Survey covered 39,063 employers across 41 countries and found 72% reporting difficulty filling roles. For the first time, AI capabilities topped the list of hardest-to-find skills, ahead of every traditional IT and engineering discipline. AI model and application development led globally at 20%, with AI literacy at 19% [1]. In Asia Pacific and the Middle East the picture is sharper still: 71% of employers report shortages, with AI model and application development at 27% and AI literacy at 26% [2]. When a shortage is global, a domestic-only search is not a smaller version of the right search. It is the wrong search.
An honest read on the market rate. Most boards discover what a capability costs by running a failed search first. Six months of rejected offers is an expensive way to learn that your band was set two years ago. Running a mandate across four markets gives you live comparative compensation data in weeks. You will sometimes find the number is lower than you feared. You will sometimes find your entire band is wrong.
Market-entry capability you cannot buy any other way. A Korean-speaking engineering lead who has shipped in Seoul understands buyer psychology, hiring norms and regulatory temperament in a way that no market-entry report will give you. Hire the person and you get the market intelligence for free.
Resilience. Concentrated teams have concentrated failure modes. One country's policy change, one competitor's aggressive hiring round, and your roadmap slips.
Borderless is not remote work, and it is not labour arbitrage
Two misreadings do most of the damage here, and both are worth killing early.
The first is that borderless means remote. It usually involves remote work, but the strategy is about where you look, not where people sit. Some of the best borderless outcomes we see end in relocation.
The second is that borderless means cheap. Deel's 2025 State of Global Hiring Report, drawn from its own platform data, is instructive on this. Yes, 82% of workers on the platform were remote as of 2024. But domestic hiring on the platform grew 104% in 2024 against 42% for cross-border hiring, and organisations showed a clear preference for talent in or near their own time zone. Half of cross-border hires in Great Britain were in the same time zone, and 41% of hires into Germany were within an hour of their employer's [3].
Most telling of all: among startups that had raised over 100 million US dollars, cross-border hiring concentrated in the UK, Canada and Germany, with software developers and AI engineers dominating [3]. Those are not low-cost markets. Those are expensive markets with deep technical populations.
So the mature pattern is the opposite of spray and pray. It is narrow, deliberate and expensive, aimed at capability that genuinely does not exist at home. Any adviser selling you borderless hiring as a cost-reduction exercise is selling you last decade's idea.
How international sourcing really works: mapping before networking
Here is the mechanic that separates a real borderless search from a wider version of the usual one.
We start by defining the population, not the shortlist. If the mandate is a VP of Engineering who has taken a model from research into production at scale, the honest first question is: how many people on earth have actually done that, and where do they work right now? Our candidate-mapping platform answers that question by enumerating companies and teams rather than by asking who we happen to know. Company by company, team by team, across every market in scope. The output is a map, and the map is usually uncomfortable, because it shows you how small the real population is.
Then we qualify against evidence rather than titles. Job titles in AI are close to meaningless across borders. A Staff Engineer in Bangalore, a Principal in Berlin and a Lead in Singapore may be doing the same work at three different levels of seniority, or three different jobs at the same one. We qualify against shipped work: what was built, at what scale, with what constraints, and what the person actually owned.
Then we sequence by feasibility, not by preference. This is the step almost everyone skips. Once the map exists, we overlay the immigration and compliance reality of every market in it before the first approach goes out. It is the difference between a shortlist and a shortlist you can actually hire.
Our consultants then run the human half. Mapping tells you who exists. It does not tell you who will move, what their spouse's career needs, or whether they will survive your board. That still takes judgement, and it still takes someone who has done it before.
The compliance layer is where borderless strategies quietly die
If you are hiring into Singapore, the rules are not a footnote. They shape who you can realistically land, and they should shape your sourcing strategy from day one.
Start with the Employment Pass. The qualifying salary begins at S$5,600 a month for most sectors and S$6,200 in financial services, rising with age to S$10,700 and S$11,800 respectively at 45 and above. From 1 January 2027 those floors rise to S$6,000 and S$6,600, reaching S$11,500 and S$12,700 at 45 and above [4]. Clearing the floor is only stage one.
Stage two is COMPASS, and a candidate needs 40 points to pass [4]. The rubric awards points across six criteria. Two are about the individual: salary against local professional benchmarks by age, and qualifications. Two are about your firm: the share of the candidate's nationality among your professional staff, and your firm's share of local professionals within its subsector. Two are bonuses, for jobs on the Shortage Occupation List and for firms meeting criteria on investment, innovation, internationalisation or workforce transformation [5].
Read that again, because it contains the single most under-appreciated fact in Singapore technology hiring. Two of the six criteria score your company, not your candidate. If 25% or more of your professional staff already share the candidate's nationality, the diversity criterion scores zero. Below 5%, it scores the full 20. And if one nationality makes up a third or more of your professional headcount, even the Shortage Occupation List bonus is halved from 20 points to 10 [5]. Smaller firms get some relief: below 25 professional employees, you score 10 points by default on both firm criteria [5].
The practical consequence is blunt. Your nationality mix is not an HR diversity statistic. It is a hiring input that determines whether your next candidate is approvable. A company that has hired its last eight engineers from one country has quietly made its ninth hire from that country much harder, and nobody tells you this until an application is rejected. A genuinely borderless sourcing strategy protects your COMPASS position. A lazy one destroys it.
Then there is the Fair Consideration Framework. Before most Employment Pass applications, the role must be advertised on MyCareersFuture for at least 14 consecutive days, and local candidates must be fairly considered [6]. Two exemptions matter in practice: companies with fewer than 10 employees, and roles paying a fixed monthly salary of S$22,500 or more, a threshold set to correspond to the top 10% of pass holders [7]. Plan the 14 days into your timeline. It is not a delay you can compress later.
At the top of the market, the Overseas Networks and Expertise Pass changes the calculus again. It requires a fixed monthly salary of at least S$30,000, runs up to five years, is not tied to a single employer, and lets the holder concurrently start, operate and work for multiple companies [8]. For a founder-operator or a portfolio technologist, that flexibility is often worth more than the salary itself, and it is a card too few boards know they can play.
Five tests before you commit to a borderless search
- Count the population before you brief anyone. If nobody can tell you roughly how many people worldwide have done the thing you need, you are not ready to start. Insist on the map.
- Check your own COMPASS position first. Pull your professional headcount by nationality and your local share. Do it before you fall in love with a candidate, not after.
- Decide what remote actually means to you. Fully distributed, time-zone-adjacent, or relocation. Say it out loud at the briefing. It changes the entire target list.
- Set the band from comparative data, not from last year's. If your band was built for a domestic market, expect it to be wrong in both directions.
- Sequence immigration before outreach. Every week you spend approaching candidates you cannot land is a week of goodwill spent for nothing.
Frequently Asked Questions
What is the difference between cross-border hiring and a borderless talent strategy?
Cross-border hiring is a transaction: you needed someone, you found them abroad. A borderless strategy is a standing position. You define roles by the capability required, then search wherever that capability exists, with the compliance and relocation machinery already in place. One is reactive. The other is designed.
Does borderless hiring only make sense for large enterprises?
No, and in Singapore smaller firms have a specific advantage. Companies with fewer than 25 professional employees score 10 points by default on both firm-related COMPASS criteria [5], and those with fewer than 10 employees are exempt from the MyCareersFuture advertising requirement [7]. Smaller organisations often move faster through the process than large ones.
Will we save money by hiring internationally?
Probably not, if you are hiring senior technical leadership. The evidence points the other way: well-funded companies concentrate their cross-border hiring in expensive markets like the UK, Canada and Germany, for expertise rather than cost [3]. Budget for capability, not for arbitrage.
How long does an Employment Pass add to a search?
Can we hire someone who will not relocate to Singapore?
Often, yes, but the employment structure changes and so does the tax and payroll position in their home country. Decide this before you build the shortlist. Reversing it after an offer is where most borderless searches lose their candidate.
The takeaway
A global network tells you who a firm knows. A borderless strategy tells you who exists. The second question is the harder one, and for the roles that actually decide whether your AI programme ships, it is the only one worth asking.
If you are weighing an international search for technology leadership across Singapore or APAC, talk to us. We will start with the map.
Sources
- manpowergroup.com: ManpowerGroup 2026 Talent Shortage Survey, 26 February 2026. Global survey of 39,063 employers across 41 countries; 72% report hiring difficulty and AI skills rank hardest to find.
- prnewswire.com: ManpowerGroup APME release, 26 February 2026. Regional figures for Asia Pacific and the Middle East, covering 12,193 employers across 10 markets.
- deel.com: Deel, State of Global Hiring Report 2025. Platform data on remote share, domestic versus cross-border hiring growth, time-zone preference, and where well-funded startups concentrate cross-border hires.
- mom.gov.sg: Singapore Ministry of Manpower, Employment Pass eligibility. Qualifying salary by sector and age, the changes taking effect 1 January 2027, and the 40-point COMPASS requirement.
- mom.gov.sg: Singapore Ministry of Manpower, COMPASS Scoring Rubric (Annex A). Full points table for criteria C1 to C6, the nationality-share thresholds, the reduced skills bonus, and the default scoring for firms with fewer than 25 professional employees.
- mom.gov.sg: Singapore Ministry of Manpower, Fair Consideration Framework. The requirement to advertise on MyCareersFuture for at least 14 consecutive days and to consider all candidates fairly.
- mom.gov.sg: Singapore Ministry of Manpower FAQ. The S$22,500 fixed monthly salary advertising exemption and its relationship to the top 10% of pass holders.
- mom.gov.sg: Singapore Ministry of Manpower, Overseas Networks and Expertise Pass key facts. Validity of up to five years, no tie to a single employer, and the right to work for multiple companies concurrently.