Global Talent #53

Brought to you by Lundi: we design, hire, and run international teams.

🔥 Opening Shot

In 2015 I was living in Guangzhou and I needed exactly one person: a multilingual Salesforce architect to work across teams in Europe, the United States, and a China office that was almost entirely expats. No Chinese required. In a country of more than a billion people, I still couldn't find her.

I found her in an adjacent time zone, in another country, and we worked together remotely for a year before she eventually joined us full time. The lesson has never left me. Population is not talent. Headcount is not capability. Being surrounded by people does not mean you are surrounded by the person you actually need.

I thought about that this week when the Philippines, one of the largest offshore services markets in the world, published its 2028 outlook and said the quiet part out loud. The head of the industry association called it a shift "from capacity to capability." Translation: the sector that built itself on how many people it could put in seats has stopped selling seats as the growth story.

That landed in the same fortnight the US Bureau of Labor Statistics reported that the seats you already have keep getting more expensive. Not through wages, which barely moved. Through the parts you never negotiate. Health benefit costs in the US rose 6 percent over the year while wages rose 3.1, and inflation-adjusted wages went backwards.

So the seat costs more and means less. Most companies are still planning in seats anyway.

Stop asking how many people you need. Ask what capability you are trying to own, and then decide where it should live.

This Week's Number: 6.0% — the 12-month rise in US private-industry health benefit costs to June 2026, roughly double the 3.1% rise in wages (BLS Employment Cost Index, released July 31, 2026).

📌 On the Radar

1. The Philippines just stopped selling headcount.
IBPAP released its 2028 industry outlook on July 14, projecting US$43.3–50.5 billion in revenue and 1.85–2.14 million full-time employees by 2028. The roadmap it replaces, set in 2022, aimed at roughly $59 billion and 2.5 million workers. The association is explicit about why: growth "will be driven less by workforce scale and increasingly by higher-value capabilities and AI-enabled talent," with global capability centers, healthcare and financial services named as the segments doing the pulling. CEO Jack Madrid put it plainly: "For years, our industry has measured success by how many people we could employ… That is the shift from capacity to capability."

When the supply side of the offshore market tells you seats are no longer the growth story, it is worth asking why so many buy-side contracts are still priced per seat. A per-head model scales with heads. If the value is moving to what each person can actually do, then the instrument you are buying is drifting away from the thing you want. This is not an argument against starting with an employment partner. Starting that way is correct, and it is how we start too. It is an argument about what comes after. Employing ten people is the easy half; recruiting the right ten, and operating them well is the half that compounds.

2. The cost of a US seat is rising in the places you do not negotiate.
The BLS Employment Cost Index for June 2026, published July 31, shows private-industry compensation up 3.3% over 12 months, with wages and salaries up 3.1% and benefit costs up 3.8%. Health benefits alone rose 6.0%. Inflation-adjusted wages fell 0.4% over the year, and civilian compensation overall rose 3.4%.

Read those two lines together and you get the whole problem. Your merit matrix is a wage document. Your cost curve is a benefits document. Health costs compounding at 6 percent will outrun anything your comp committee is willing to approve, and they attach to a person rather than to performance. Meanwhile your people feel poorer, because in real terms they are. That gap is where the "we cannot afford this team" conversation starts, and it usually ends in a headcount freeze. The honest read is different: the loaded cost of a seat in one geography is climbing for reasons nobody in your building controls. That is an org-design problem wearing a payroll costume.

3. The UK's compliance bill arrives in October. Then again in January.
The government's implementation timeline sets out what lands next. From October 2026, UK employers must take "all reasonable steps" to prevent sexual harassment, become liable for harassment of their staff by third parties, must tell workers about their right to join a union, and face strengthened union access rights. From January 2027, the unfair dismissal qualifying period drops from two years to six months and compensatory awards are uncapped, though the government notes that 2027 timings remain under review.

None of this is a reason to avoid the UK. It is a reason to stop treating employment risk as something that just happens to you. Every one of these obligations has an owner, a process, and a cost, and none of it shows up on a rate card. The practical question for a COO is which roles genuinely need to carry UK employment risk and which are better built somewhere else, with a partner who operates that compliance layer as part of the job rather than handing it back to you. Whoever holds the employment relationship holds the liability. Make that a decision rather than a default.

📊 Chart of the Week

Caption: US private industry, 12-month change in employer costs for the year to June 2026. Source: BLS Employment Cost Index, released July 31, 2026.

Wages rose 3.1% and lost to inflation. Benefits rose 3.8%. Health benefits rose 6.0%. The line your compensation committee argues over is the smallest, slowest-moving piece of the number, and the fastest-moving piece is the one nobody in the room can negotiate. This is why "cut the merit budget" never fixes anything: you are trimming the part that grows slowly to offset the part that grows quickly. The only lever with real range left is the one most companies treat as fixed, which is where the work sits and who operates it.

🚀 One More Thing

If your 2027 plan is still a number of seats, it is already out of date. Book a strategy session and we will go through it role by role: what capability you are actually trying to own, which of it truly has to sit where it sits, and what the fully loaded number looks like once benefits and compliance are in the picture. No pitch. Just insight from someone who has been there.

📖 New here? My book, Winning the Global Talent War, is the full playbook behind this newsletter.