A family from Shenzhen messaged me last month: they had spent SGD 14,000 on six months of one-on-one AEIS tutoring for their Primary 4 son. He still didn't pass. Another family, same age, same starting English level — spent under SGD 1,500 on a mix of self-study books, an online platform, and one human tutor for the last four weeks. He passed. The difference wasn't luck. It was how they thought about AEIS tutor ROI before paying anyone a single dollar.
What "ROI" Actually Means for AEIS Prep
Return on investment, when you strip away the spreadsheet language, is just this: for every dollar and every hour you put in, how much closer does your child get to a pass letter from MOE?
Most overseas parents calculate this badly. They look at a tutor's hourly rate and either flinch or shrug. Neither reaction is useful. The real ROI equation has four variables:
- Cost: tuition fees, materials, platform subscriptions, opportunity cost of parent time
- Time-to-readiness: weeks until the child can walk into the test and finish under timed conditions
- Probability of passing: realistic, not the tuition centre's marketing promise
- Downside protection: what happens if the child fails — do you have data to debug?
A SGD 200/hour tutor who pushes pass probability from 35% to 75% is cheaper, per percentage point gained, than a SGD 60/hour tutor who pushes you from 35% to 45%. This is the lens I want you to keep using through the rest of this article.
AEIS Tutor Cost: The Real 2026 Numbers
Let me lay out what I actually see families pay in Singapore in 2026. These are ranges, not promises — every tutor and centre prices differently.
| Format | Hourly / Session Cost (SGD) | Typical Total for 4-6 Months |
|---|---|---|
| Centre group class (6-12 students) | 50-90 per 2hr session | 2,400-5,000 |
| Small group (3-5 students) | 80-130 per 2hr session | 4,000-7,500 |
| Private 1-on-1 with experienced AEIS tutor | 100-180 per hour | 8,000-18,000 |
| Ex-MOE / specialist 1-on-1 | 180-300 per hour | 15,000-30,000+ |
| Online platform / self-study app | 20-100 per month | 200-1,000 |
| AI-based prep (adaptive practice) | 30-150 per month | 300-1,500 |
A typical Beijing or Jakarta family I speak with budgets between SGD 6,000 and SGD 12,000 for the AEIS prep window. That's not small money. And it sits on top of the relocation, housing deposit, and the international school fees you're already paying while waiting for placement — which I broke down in Singapore MOE School Fees Breakdown: Real Costs for 2026.
Here's the part centres won't tell you: the marginal hour matters more than the total hours. The first 40 hours of structured AEIS prep do roughly 70% of the work — fixing major gaps in vocabulary, model drawing, and exam timing. Hours 40 through 120 are where most of the money disappears for diminishing returns.
When a Private Tutor Genuinely Earns Its Fee
I'm not anti-tutor. I'm anti-default-tutor. There are four situations where private 1-on-1 is the highest-ROI choice:
1. The diagnostic gap is large and weird. If your child reads English fluently but can't write a coherent paragraph, or solves Olympiad-level math at home but freezes on Singapore-style word problems, you have a translation problem, not a knowledge problem. A human tutor can identify the mismatch in 90 minutes. An app cannot.
2. The child is 10 or under and needs accountability. Younger kids rarely self-direct through a 4-month prep cycle. A tutor who shows up twice a week creates a forcing function that no platform login screen replicates.
3. You're inside the final 6-8 weeks before the test. This is the highest-ROI window for human help. Mock test review, error pattern analysis, and exam-day pacing strategy are conversations, not videos. Spending SGD 3,000-5,000 in the final stretch often outperforms SGD 8,000 spread over six months.
4. English is the bottleneck, especially comprehension. Reading inference questions need a coach who can model thinking out loud. I wrote about why in AEIS English Comprehension Strategies That Actually Work in 2026 — the techniques are coachable, but only by someone who can hear how your child reasons.
Outside these four, the ROI math gets shaky fast.
AEIS Prep Online vs Tutor: A Cleaner Comparison
The "online vs tutor" debate is framed wrong. It's not either/or — it's about sequencing. Here's how I'd actually structure 6 months of prep for a Primary 3-5 candidate:
| Phase | Months | Primary Mode | Why |
|---|---|---|---|
| Diagnostic | Month 1 | 2-3 hours with a tutor + one online mock | Identify gaps fast |
| Foundation | Months 2-3 | Online platform + self-study books | Cheap, scalable, repeatable |
| Skill building | Month 4 | Hybrid — online drills + 1 tutor session/week | Targeted intervention |
| Exam simulation | Months 5-6 | Tutor 2x/week + weekly timed mocks | Pacing, error analysis |
Run this way, total spend lands around SGD 4,500-7,000 — roughly half of a tutor-only path, with comparable or better outcomes. The reason is simple: drilling vocabulary lists, practicing arithmetic, and watching grammar explanation videos do not require a SGD 150/hour human. Coaching error patterns and decoding exam psychology do.
Families who default to tutor-only often pay tutors to babysit drill work. That's the single biggest source of wasted AEIS prep budget I see.
AEIS AI vs Tutor: What 2026 Tools Actually Do Well
AI-based prep has changed meaningfully in the last 18 months. I run an AEIS prep platform myself, so treat what follows with appropriate skepticism — but also know I see thousands of practice attempts a week, which informs what I'm about to say.
Where AI clearly outperforms a human tutor:
- Volume of practice generation: a child can drill 50 model-drawing variations in an hour. No tutor produces that many fresh problems.
- Instant feedback on objective items: vocabulary, grammar, computation. A human marking the same paper takes 24 hours and costs SGD 80.
- Adaptive difficulty: a good system raises difficulty when the child is ready and backs off when they aren't, in real time.
- Cost per practice hour: roughly SGD 1-3 versus SGD 80-180 for a tutor.
Where AI is still meaningfully worse:
- Open-ended writing feedback: improving, but not yet as good as a thoughtful English teacher reading two compositions side by side.
- Reading the child's emotional state: a tutor sees the kid is shutting down and changes tactic. AI doesn't.
- Subtle reasoning errors in word problems: AI can mark the answer wrong; a tutor can see why the child went wrong.
- Parent communication and exam strategy: humans win.
The honest framing: in 2026, AI replaces about 60-70% of what a mid-tier tutor used to do, at maybe 5% of the cost. It doesn't replace the top 30% — diagnostic judgment, motivation, exam-day coaching. So the AEIS AI vs tutor question shouldn't be a fight. The right answer is: use AI for the 70% it does well, pay a human for the 30% it can't.
A Real ROI Calculation Worked Out
Let me walk through a hypothetical but typical family. Mainland Chinese, son aged 11, target Primary 5 placement, current English at roughly Singapore Primary 3 level, math at Primary 5+ level.
Path A: Tutor-only (1-on-1, twice a week, 5 months)
- 40 sessions × 1.5 hours × SGD 150 = SGD 9,000
- Materials and mock tests: SGD 400
- Total: SGD 9,400
- Estimated pass probability: 65-75%
- Cost per percentage-point gained (from baseline 30%): roughly SGD 235
Path B: Hybrid (AI platform + targeted tutor)
- AI platform 5 months: SGD 500
- Self-study books and past papers: SGD 200
- 12 tutor sessions in final 8 weeks × 1.5 hours × SGD 150 = SGD 2,700
- Mock tests + diagnostic: SGD 300
- Total: SGD 3,700
- Estimated pass probability: 65-75% (same range)
- Cost per percentage-point gained: roughly SGD 92
Path C: AI / online only
- Platform + materials: SGD 1,200
- Total: SGD 1,200
- Estimated pass probability: 45-55% (depends heavily on parent involvement)
- Cost per percentage-point gained: roughly SGD 60, but absolute pass rate is lower
Path B is the highest-ROI choice for most families. Path C is rational only if budget is genuinely tight or the child is unusually self-directed and the parent is hands-on. Path A makes sense for families with abundant budget and zero appetite to coordinate logistics — you're effectively buying convenience at SGD 5,000.
The Hidden Costs Nobody Lists
Tutor cost is the line item parents fixate on. The hidden costs are larger:
- Wrong-tutor risk: 3 months with a tutor who teaches Singapore math wrongly costs you SGD 4,000 and three months. Verify they actually understand the model method — see AEIS Math Model Drawing Method: A Pillar Guide for Overseas Parents for what competent looks like.
- Logistics tax: if you're not in Singapore yet, scheduling tutors across time zones eats parent hours. Build that into the budget.
- Document and registration overhead: I covered this in AEIS Registration Document Checklist 2026: What Overseas Families Actually Need — missing a deadline is the worst possible ROI outcome, because it kills the entire investment cycle.
- Opportunity cost of the child's time: 200 hours of tutor sessions is 200 hours not spent on Chinese/Bahasa, sports, or sleep. Sleep loss in particular compounds badly in the final month.
How to Actually Decide
A simple decision sequence:
- Get a real diagnostic first. Before you commit to any package, pay for one or two diagnostic sessions or take a serious mock. Know your starting point in numbers, not feelings.
- Set a budget ceiling and timeline. "I will spend up to SGD X across Y months" — write it down. This stops emotional escalation when results in month 2 look uneven.
- Default to hybrid. Online or AI for volume, human for judgment, weighted toward the final 8 weeks.
- Re-evaluate at month 2 and month 4. If the trajectory is flat, change something — usually the tutor, sometimes the child's study schedule, occasionally the target level.
- Plan for two attempts. AEIS allows two attempts at most. Budget and emotional energy should assume the first might miss. The full timing context is in AEIS Complete Guide 2026: Timeline, Test, and Real Costs, and the relocation context in Moving to Singapore with School-Age Kids — A Checklist.
Next Steps
- Run a baseline mock this week and write down two numbers: English score and Math score. Without these, every spending decision is guesswork.
- Set a hard prep budget. SGD 4,000-7,000 is enough for most families if structured properly. Anything above SGD 12,000 needs a specific reason.
- Sequence your spending: cheap and high-volume first (AI / platform / books), expensive and judgment-heavy last (tutor in final 6-8 weeks).
- Pick one tutor candidate and book a single trial session before any package. The trial-to-package conversion is where most overpayment happens — slow it down.