Every year the sector produces a fresh set of headline statistics. ISC Research’s Global International Schools Snapshot 2026 counts 15,075 international schools worldwide, teaching 7.7 million students, employing 730,000 staff and generating around USD 69.3 billion in annual fee income.
Those figures are useful for framing a market. They tell an investor almost nothing about whether a specific project will work.
The numbers that decide that question are smaller, less quotable and rarely published. They sit inside financial models, staffing plans and enrolment forecasts. Since 2011 GSE has delivered more than 60 school and education projects across 28 countries, almost all of them from nothing. Across three articles, these are the fifteen figures worth arguing about. This first one covers the four that are settled before a single student enrols, because they are the four that cannot be changed afterwards.
A note on how to read them. Each carries a line stating where it comes from: a published figure with its source named, a figure derived from data GSE already publishes, or a practitioner range drawn from live projects. No figure travels above its basis. Every one of them moves with market, curriculum, fee position and land cost. Treat them as the starting point for a conversation with your model, and be suspicious of anyone who quotes a single figure without asking what kind of school you are building.
1. Capital cost per seat
Derive it, do not quote it.
Take the published construction cost per square metre for a four-star hotel in the target city, multiply by 10 to 12 square metres of gross floor area per student, and add roughly a quarter for fit-out, furniture, external works and fees. Land sits outside the figure.
A hotel and a school are cost cousins: mid-range finishes, large public rooms, kitchens, dense mechanical and electrical services, and a fit-out that has to survive hard use. The hotel figure is published every year by quantity surveyors for most cities an investor is looking at. A school-specific figure usually is not.
That is a first-pass screen, not a cost plan. GSE uses it to test whether a site, a fee level and a specification belong in the same sentence before anyone spends money on design. The hotel line comes from a current quantity surveyor publication for the region; Turner and Townsend’s market intelligence reports publish it by grade for the Gulf and most Asian and African cities, and a four-star build sits between their midscale and luxury tiers. A project that fails at this level is not rescued by refinement. One that passes goes on to a proper cost model at feasibility, which replaces every number in the screen.
Worked example. Turner and Townsend’s 2025 UAE report puts a midscale hotel in Dubai at roughly USD 1,770 per square metre of direct build cost. At 11 square metres per student and a quarter on top, that is about USD 24,000 per place before land. Push the specification towards four-star, at roughly USD 2,200 per square metre, and the figure approaches USD 30,000. For scale, GEMS has said its new Dubai Sports City campus cost USD 100 million on 47,600 square metres, which is about USD 2,100 per square metre of gross area for a super-premium build, and 30 per cent more than its previous premium projects. Riyadh runs materially higher than Dubai in the same survey; Jakarta runs at about half.
The spread between cities is enormous, and it is where feasibility studies quietly fail. The usual error is choosing a facility specification before the fee level has been tested, which is the sequence a feasibility study exists to enforce. Specification should follow the fee the market will pay, and the fee should follow honest demand analysis. Schools built in the other order carry the specification for twenty five years.
Basis. Method. The area allowance is published in GSE’s article on the area required for a school and is consistent with UK Building Bulletin 103 gross area norms for secondary schools. The hotel cost lines are published by Turner and Townsend (UAE and KSA Market Intelligence 2025). Arcadis’s International Construction Costs report is a relative index and does not carry a per square metre hotel line. The GEMS campus cost is as stated by GEMS, reported by Khaleej Times, 2025. The quarter for fit-out and fees is a GSE practitioner allowance.
2. Fee level against operating cost
At maturity, a well-run school spends 65 to 75 per cent of fee income on operations before rent, debt service or return.
That relationship governs everything else in this list. Staff take 45 to 55 points of it. Everything else the school consumes, from utilities and maintenance to learning resources, marketing, insurance and administration, takes the other 15 to 25. What remains is the margin the property and the investor have to share. If the achievable fee in your market is USD 6,000 and operating cost per student lands at USD 4,500, there is very little left once the building is paid for. Investors who fall in love with a premium facility in a mid-fee market usually discover this after the concrete has been poured.
Do not test that figure against a listed group’s accounts without adjusting. Nord Anglia, Curro and ADvTECH all report group EBITDA margins in the low to mid twenties, which implies operating costs of 75 to 78 per cent of revenue. Those figures are struck after property costs and blend mature schools with ramping ones. A single mature school before rent runs leaner than the group that contains it.
Basis. GSE practitioner range for a single mature school before rent. Group operating cost ratios of 75 to 78 per cent after property from Nord Anglia (F-1, 2014, and FY2016 results), Curro (Annual Integrated Report 2024) and ADvTECH (2025 results), all party-published.
3. Time from agreement to first student
Realistic range: 18 to 36 months for a new build. Nine to fifteen months for a fit-out or conversion.
Regulatory approval is the variable that breaks schedules, and it varies wildly by jurisdiction. Saudi Arabia, the UAE, Indonesia and Vietnam each run different licensing sequences with different sticking points. Build the licensing critical path before the construction programme, because a school that misses an August opening loses a full year of revenue and a full cohort of families.
The fast end can be beaten where a government or a developer delivers the shell. Nord Anglia was allocated its Kowloon site by the Hong Kong Education Bureau in April 2013 and opened in September 2014, about seventeen months, and that case is the exception. In most jurisdictions the constraint is the licence rather than the construction.
Basis. GSE practitioner range, from projects delivered since 2011. The Hong Kong timeline is party-published in Nord Anglia’s F-1, 2014.
4. Total funding requirement to break-even
Usually 25 to 40 per cent above the construction cost.
This is the number investors underestimate most often. The build is usually financed. The three to five years of operating losses that follow it, sometimes longer, often are not funded at all. A school that opens with only its capital budget funded starts cutting staff and marketing in exactly the years when it can least afford to.
No operator publishes this figure, so treat it as a modelling assumption rather than a benchmark. The evidence for the shape of it is in the accounts of individual start-ups: Nord Anglia’s Dublin school, opened in 2018, was still loss-making in its sixth year according to its Irish company filings, with staff costs rising a fifth in a single year as the roll grew. The uplift over construction cost is the working capital that carries a school through years like those.
Basis. GSE estimate. The break-even timeline is consistent with GSE’s published guidance that most international schools take three to seven years to reach stable enrolment. The Dublin loss profile is from Irish company filings as reported in April 2024.
What to test
Before anything else in a project, test one relationship: does the fee level support the facility specification, with room left for rent and return? If it does not, nothing in the enrolment forecast or the staffing plan will rescue it. If it does, the next four figures decide whether the school fills. They are the subject of the second article.
Read next
- International Education by the Numbers, Part Two. The four figures that decide whether the school fills.
- International Education by the Numbers, Part Three. The seven figures that decide whether it lasts.
- The Real Cost of Starting an International School. Where the capital goes, and the three to seven years to break-even.
- How to Conduct a Feasibility Study for a New School. The document that tests fee, specification and demand in the right order.
- Why Most New School Projects Fail Financially. The assumptions that survive a model and not a school.
To talk about your school project, reach us through the contact page.