Optimizing People Costs Starts With Getting Headcount Right

Key Takeaways

• Headcount planning is not a counting exercise. It is how a company designs its labor cost structure, which typically accounts for 50 to 70% of operating expenses.
• The true cost of an employee runs well above salary. Mandatory insurance alone adds 21.5% on top of pay in Vietnam, before recruitment, onboarding, and a replacement cost that can reach 200% of annual salary for senior roles.
• 47.7% of companies plan to hold headcount steady for 2026, but freezing headcount does not equal saving money without a genuine optimization step.
• AI is lowering the number of people needed for the same volume of work, so companies that update their productivity assumptions can optimize cost without mechanical cuts.

With salary increases at a decade low and AI reshaping demand for labor, headcount planning is shifting from gut feel to a data exercise. According to the Talentnet-Mercer 2025 Salary Survey, 47.7% of companies plan to hold headcount steady for 2026. This article moves from how to calculate headcount, to the real cost of an employee, to how to optimize that cost without mechanical cuts.

Headcount planning is more than counting heads

Headcount planning is the process of defining the number, structure, and quality of people a business needs to operate effectively against its goals. Unlike the common view, it is not a static figure of names on a payroll. It is a strategic decision about cost and capability.

Three terms are often used interchangeably but measure different things.

For an executive team, headcount deserves attention because people cost typically runs at 50 to 70% of operating expense. A weak plan creates two symmetrical costs. Overstaffing lets labor cost erode margin while output per employee drops below benchmark. Understaffing pushes teams into overload and overtime, service quality slips, attrition rises, and the result is a spiral of being ever more short-handed.

How to calculate headcount through formulas and process

A credible headcount figure does not come from intuition. It comes from three layers that build on each other: a foundational formula, a workload-based model, and a revenue-linked model.

The foundational headcount formula

The most basic formula divides the total work hours required by the available work hours per person, then adds a buffer for expected attrition and a growth factor. Two metrics belong alongside it and deserve regular executive attention: revenue per employee, calculated as total revenue divided by total headcount, and labor cost as a percentage of revenue. Together they show how much value the organization generates for every dong spent on people.

FTE models by workload and by revenue

For operations-heavy functions such as manufacturing, contact centers, or logistics, a workload-based FTE model is more accurate. As an illustration, 40,000 cases a month at an average handling time of 420 seconds equals roughly 4,666 work hours. After subtracting schedule adherence, shrinkage, and a realistic utilization rate, each person delivers about 105 useful hours a month, which implies a need for roughly 44 people.

For an MNC planning across several functions, a revenue-based FTE model ties headcount directly to business results. Forecast revenue is allocated to each function, a productivity improvement assumption is applied, and the required FTEs are derived from a target revenue-per-FTE figure. This approach holds each function accountable for the revenue it generates and curbs the tendency to expand the team at a department head’s discretion.

A five-step process for Vietnam

A formula is only one link. To make headcount a repeatable discipline rather than strategic staffing done ad hoc, companies in Vietnam commonly apply five steps: define labor needs against business goals; analyze the current workforce; calculate requirements with the formulas; build the staffing plan with a budget; and monitor and adjust against attrition and market shifts. Talentnet’s fuller framework extends this to seven steps, adding a scenario-building stage and the tracking of metrics such as revenue per employee.

The real cost of an employee that leaders underestimate

Most headcount decisions stop at the salary figure, but the true cost of an employee is far higher. In Vietnam, the full labor cost has three parts: salary and allowances, mandatory contributions, and recruitment cost.

On mandatory insurance alone, an employer pays an additional 21.5% on top of salary for Vietnamese employees. The 2% trade union fund is a separate item, which lifts the employer’s total statutory obligation to roughly 23.5%. An employee on VND 35 million a month therefore carries a total cost of about VND 43 million, before any bonus, 13th-month pay, or health benefits.

The least-counted expense is the cost of replacing an employee. Replacing a person can cost 50 to 200% of annual salary depending on level, reaching around 200% for leadership roles (Gallup), while SHRM estimates the equivalent of 6 to 9 months of pay, not counting the knowledge and productivity lost. Against that, voluntary turnover in 2025 ran at 6.6% in MNCs and 9.8% in local companies (Talentnet-Mercer 2025).

The current context makes the cost question more urgent. According to Talentnet-Mercer 2025, 47.7% of companies plan to hold headcount steady for 2026, while average salary increases hit a decade low. But holding headcount steady is not the same as holding cost steady. A frozen structure can still hide productivity gaps, internal pay inequities, and roles that no longer generate proportional value.

Optimizing headcount through reallocation, not cuts

Effective optimization does not start with the question of how many people to cut. It starts with which work needs people and which can be reallocated. The four moves below reinforce one another.

Segmenting the hard core and soft core

The first move is to segment the workforce into two layers. The hard core holds the critical, hard-to-replace roles that must be retained at all costs, developed deeply, and paid competitively. The soft core holds support and project-based roles managed flexibly through contracts, outsourcing, or seasonal staff, able to scale with demand. Done right, this converts part of fixed cost into variable cost without weakening operating capacity. As Ms. Nguyen Thi Quynh Phuong, Head of Human Capital Solutions at Talentnet, puts it, companies need to clearly define their hard core as the stable workforce they must keep and their soft core as the flexible workforce they can adjust as needed, and that flexibility must live not only in HR policy but in leadership mindset.

Balancing the Build, Buy, Borrow, Bot model

The second move is to balance four ways of meeting capability needs, known as the 4B model. Build means developing people internally. Buy means hiring externally for critical, hard-to-fill roles. Borrow means using secondments, advisors, or outsourcing for short-term needs. Bot means using AI and automation for repetitive, rules-based work.

Of the four, Bot acts most directly on the headcount number. Once part of the work is automated, the assumption of how much one person can handle has to be recalculated, which usually pulls required headcount below what an old FTE model, still built on pre-AI productivity, would suggest. MISA is one example: by bringing AI into its operations, it reduced its workforce from 600 to 350 people without a proportional drop in output. The point is not to remove people but to redeploy them away from repetitive work toward value creation. At the other end, FPT leans on Build, pairing competitive pay with a learning culture to lower the real cost of each employee it retains.

Planning across three scenarios

The third move is to plan across three scenarios rather than a single number. The upside case allows faster hiring and an expanded flexible layer. The base case holds headcount and develops from within. The downside case relies on natural attrition and a backfill freeze rather than layoffs. With MNC voluntary turnover at around 6.6%, a target of cutting headcount by 5% can be met within 12 months through a backfill freeze alone, without a single layoff decision. At larger scale, planning ahead is decisive: workforce forecasting and early sourcing let major projects fill thousands of roles on schedule instead of scrambling to hire at a premium.

Conclusion

Optimizing headcount is ultimately about structure, not numbers: a smaller team of the right people outperforms a larger team that is misaligned.

Sustainable growth therefore comes from having the right people in the right roles at the right cost, not from how many people there are. The move for a leadership team is to treat headcount as a strategic decision about cost and capability, calculated on data rather than instinct, and optimized through reallocation rather than mechanical cuts. Talentnet’s HR Consulting and Organizational Development services, together with Talentnet-Mercer salary survey data covering 678 organizations, help companies calculate, benchmark, and optimize headcount from the planning stage.