# Monday Briefing — 22 June 2026
**Simply English | Weekly Ops & Strategy**

---

## 📊 This Week's Focus

Three key themes coming out of last week's work:
1. **AI marking opportunity** — competitive landscape fully mapped, technical plan ready
2. **Satellite model expansion** — space efficiency under current footprint
3. **Competitive watch** — LingoTask's model is instructive but not a direct threat at centre level

---

## 1. 🏢 Space & Operational Efficiency

### Current State
- 3 learning centres (Simply English and Learning) in HK
- Satellite teaching model already in play — offsite or shared-space classes

### Observations
- The LingoTask research highlighted one critical metric: **AI marking saves teachers ~70% grading time**
- For your centres, this translates to:
  - Same teacher → more student throughput (without adding desks)
  - Less after-hours marking → better teacher retention
  - Faster feedback cycle → higher perceived value for parents

### Space Efficiency Takeaway
AI marking doesn't require more physical space, but it **unlocks capacity** within existing space. If a teacher spends 10–15 min per essay manually, and AI drops that to 2 min review, you effectively free up ~80% of grading time for teaching or additional students.

---

## 2. 🏆 Competitive Landscape

### Direct Competitors in HK Learning Centre Space

| Competitor | Threat Level | Notes |
|---|---|---|
| **Monkey Tree** | Low–Med | No AI marking observed; traditional model |
| **Headstart** | Med | Could adopt LingoTask or similar and gain efficiency edge |
| **British Council** | Low | Well-resourced but different market segment |
| **LingoTask** | **Med** (indirect) | School-focused (HK$29K–49.5K/yr), not learning centres — BUT if they pivot to centre pricing, that's a problem |
| **Independent tutors** | Low individually | But AI tools make them more competitive per-hour |

### The AI Gap
LingoTask's analysis showed a clear **strategic vulnerability**: none of the major HK tutoring chains seem to have an in-house AI marking solution. If your competitors adopt one before you do, they gain a significant cost/efficiency advantage. **First-mover advantage in the learning centre space is real.**

---

## 3. 🛠️ Products — What's Ready to Build

### MVP: AI Essay Marking Tool
**Status:** Technical plan completed (20 June). Ready for build decision.

| Feature | Ready? | What's Needed |
|---|---|---|
| Vision-based grading | ✅ Plan written | Hire freelancer (HK$30–60K, 4–6 wks) |
| DSE/IELTS rubric alignment | ⏳ Need your rubrics | Collect 20-30 graded samples |
| Teacher review dashboard | 🟢 Prompt templates ready | Design your preferred workflow |
| Speaking assessment (V2) | 🧠 Speaking module planned | Uses Whisper + LLM — cheap to add |

### Key Decision This Week
> **Do you want to build it yourself or partner with an existing tool?**
> - Build: HK$30–60K dev cost → own product → sell to other centres
> - Partner: Cheaper upfront → but zero ownership

---

## 4. 🧭 Strategy — Two Parallel Tracks

### Track A: Simply English Core Operations
- Continue satellite expansion
- Start collecting graded student essays (to build the AI corpus — needed whether you build or partner)
- Identify which centre could pilot AI marking first (suggestion: the one with highest essay volume)

### Track B: $10K/Month Side Ventures
_Reminder per 13 June: this is NOT a Simply English revenue target — it's personal side projects_
- **AI marking tool** → sell to other learning centres = recurring revenue stream ($)
- **DSE speaking prep** → standalone product (audio upload + AI evaluation) — very low build cost
- **Writing feedback service** → branded SaaS for HK learning centres at HK$2K/mth/centre

---

## 5. 🚀 Expansion Planning

### Short-term (Next 60 Days)
1. **Collect 20–30 graded essay samples** from any Simply English class — this is the only blocker on AI marking
2. **Decide build vs partner** for AI marking
3. **Test one AI-graded assignment** with a single class (even if through a raw LLM prompt + manual review — prove the concept)

### Medium-term (3–6 Months)
4. **Launch MVP AI marking** within Simply English
5. **Package as product** for other HK learning centres
6. **Add speaking assessment** (Whisper-based — ~2 weeks dev)

### Long-term (6–12 Months)
7. **Multi-centre SaaS** (50+ centres → ~$150–200/mth API costs, HK$2K/mth per centre revenue)
8. **Potential to expand subjects** (not just English — Chinese marking is the highest-demand gap)

---

## 6. ⚙️ Operational Improvements This Week

| Action | Priority | Owner |
|---|---|---|
| **Collect 20–30 graded essays** (any subject, any level) | High | Centre heads |
| **Decide build vs partner** for AI marking | High | Roger |
| **Define DSE/IELTS rubric you want AI to follow** | Medium | Roger + teachers |
| **Review AI marking MVP plan** (file dated 20 June) | Medium | Roger |
| **Talk to 1 centre about piloting AI grading** | Low | Roger |

---

## 🎯 Key Numbers This Week

| Metric | Current Status | Target |
|---|---|---|
| AI marking build timeline | Plan complete | Decision this week |
| API cost for Simply English volume | ~$3–5/mth | Negligible |
| Freelancer dev cost | HK$30–60K (one-time) | Vs HK$2K/mth/centre revenue |
| Competitor AI adoption | None spotted yet | Watch window: 6–12 months |

---

## 📌 Bottom Line

> **The AI marking piece is ready to act on. Last week mapped the full competitive landscape and produced a buildable technical plan. The only remaining input is your graded essay samples and a build/partner decision.**

> Meanwhile, your 3-centre operation is solid — the efficiency gains from AI marking would compound your satellite model naturally, and the first-mover advantage in the learning centre AI space is worth grabbing.

**Ready when you are on next steps.** ✌️