Every figure in the proposal, traced to a named source and a year. And — just as important — an honest register of what could not be verified, so that no decision rests on a number that isn't real.
Most market decks in this category recycle statistics that dissolve the moment anyone checks them. This dossier does the opposite: each figure carries a source and a year, contested figures are shown as contested, and unverifiable claims are named and excluded rather than quietly rounded into the argument.
Four verified facts that link into a single, defensible argument. If any one of them were false, the business case would weaken. None of them are.
79.31% of health spending is out-of-pocket (2023), and it is rising while India, Pakistan, Nepal and Sri Lanka all trend down. Verified directly from the World Bank API (indicator SH.XPD.OOPC.CH.ZS; raw value 79.30711365).
27.52% of out-of-pocket spend is diagnostic services — the second-largest category after medicines (54.40%). Consultation is only 10.31%. This single number is the TAM narrative: the money is in tests, not appointments.
Diagnostic centres pay referring doctors 20–50% commissions. The same test varies up to 32× in price between facilities. And ~95% of private facilities operate on un-renewed licences.
72.4% household smartphone ownership, 89 million active mobile-money accounts, Tk 1.71 trillion moving through MFS every month. Checkout is a solved problem. Trust is not.
The president of the Bangladesh Medical Association, Mustafa Jalal Mohiuddin, has publicly called referral commissions "a crime."
When the doctors' own professional body concedes the practice is indefensible, the reform argument stops being an outsider's complaint and becomes an insider's admission. Use it.
| Indicator | 2022 | 2023 | Source |
|---|---|---|---|
| Current health expenditure (% of GDP) | 2.22% | 2.17% | World Bank / WHO GHED |
| Health expenditure per capita | US$56.62 | US$53.46 | World Bank / WHO GHED |
| Out-of-pocket (% of health spend) | 78.25% | 79.31% | World Bank / WHO GHED |
| Derived total health expenditure: ~US$9–9.5bn (2023) — calculated from GDP × 2.17%, cross-checked per-capita. Labelled as derived, not sourced. | |||
Per-capita health spending in USD fell from $56.62 (2022) to $53.46 (2023), and its share of GDP fell too. Much of that is taka depreciation rather than a real-terms collapse — but it means you cannot honestly claim "rapidly growing health expenditure" from WHO data. Anyone who does is either not checking or hoping you won't.
The growth story is real, but it lives in two specific places: the private/out-of-pocket segment, and the public budget — which doubled to Tk 69,409 crore (1.02% of GDP) in FY2026-27, crossing 1% for the first time.
Two official OOP figures exist and a sharp reader will catch the discrepancy. WHO/World Bank says 79.31% (2023); Bangladesh's own National Health Accounts says 73% (2021 data). Both are real — different methodologies. Cite WHO/World Bank for international credibility, and cite it precisely.
The supply side is a mom-and-pop industry with almost no licensing discipline. That is precisely the profile a trusted aggregation layer can consolidate.
| What | Number | Source |
|---|---|---|
| Registered physicians | 134,568 — of whom ~36,000 practise without renewed registration (roughly 1 in 4) | BMDC, Nov 2024 |
| Doctor density | 5.26 per 10,000 — second-lowest in South Asia. Doctors+nurses+midwives: 9.9 vs the WHO SDG threshold of 44.5 | WHO |
| Diagnostic centres | 10,291 (BBS survey of operating institutions) — or 35,597 (DGHS registrations) | BBS 2021 / DGHS 2025 |
| Private hospitals & clinics | 4,452 + 1,397 (BBS) — or 19,627 (DGHS) | BBS 2021 / DGHS 2025 |
| Share of hospital beds in private hands | ~67% (105,183 private vs 52,807 public) | BBS, 2021 |
| Ownership structure | 44% sole proprietor · 40% partnership — a fragmented, capital-poor supply side with no bargaining power and, critically, idle capacity to sell. | |
The two facility datasets conflict badly and must never be averaged. BBS is a survey of operating institutions (fieldwork c. 2019–21). DGHS is a registration database (2025) that likely includes defunct and duplicate entries. Present both, name the difference, and let the regulatory chaos itself make the argument for a verified discovery layer.
Only 914 of 19,627 (4.66%) renewed their licences in the fiscal year to 27 April.
Only 1,790 of 35,597 (~5%) renewed.
~95% of private healthcare facilities in Bangladesh are operating on un-renewed licences (DGHS, 2025), and DGHS lacks the manpower to inspect them. Enforcement drives are live and ongoing through 2026 — 37 facilities sealed in a single day; ~1,000 unregistered centres shut.
Implication for the product: a valid DGHS licence must be a hard gate for listing. Listing an unlicensed centre and taking money for bookings there is the fastest route to being named in a DGHS raid.
This is the clearest product-shaped hole in the market. There is no mandated price list, no published tariff, and no comparison mechanism. As one headline put it, private diagnostic facilities "charge at will."
| Test | Government | Trust-run | Private | Max spread |
|---|---|---|---|---|
| Lower abdomen ultrasound | Tk 110 | — | Tk 1,200–3,500 | 11×–32× |
| Serum creatinine | Tk 60 | Tk 220 | Tk 500 | 8.3× |
| Lipid profile | Tk 300 | Tk 850 | Tk 1,650 | 5.5× |
| RBS | Tk 60 | Tk 170 | Tk 320 | 5.3× |
| ECG | Tk 80 | Tk 230 | Tk 400 | 5× |
| Brain MRI | Tk 3,000 | — | Tk 11,500 | 3.8× |
| Brain CT scan | Tk 2,000 | — | Tk 7,500 | 3.75× |
| CBC | Tk 150 | Tk 300 | Tk 400 | 2.7× |
| Private facilities charge up to ~20× government rates on average, and as much as 32× on individual tests. Colonoscopy runs Tk 2,500 (government) to Tk 25,000+ (private) — a 10× spread. | ||||
Source: New Age, 12 August 2022. The prices are now ~4 years stale — but the ratios are the story, and nothing has happened since to compress them. Do not present these as current retail prices; present them as evidence of structural dispersion.
A whole cottage industry of "test price in Bangladesh" SEO websites exists — low-quality aggregators competing for search traffic on diagnostic pricing queries.
That industry exists because patients are actively searching for price transparency and being served junk. The demand is already expressed, already measurable in search volume, and currently unmet by anyone credible. That is the wedge — and it is the reason the Google Search campaign in the growth plan is built on price-comparison intent rather than brand awareness.
The national emergency number exists — but it is a call-routing desk, not an EMS.
Price is negotiated at the point of maximum distress.
This is not a rhetorical risk. Ambulance syndicates control hospital gates, block outside ambulances, and take roughly half the fare as a cut from any vehicle from another district.
Strategic consequence: an app that dispatches an outside ambulance to a hospital gate will collide with these people. The answer is not to fight for the gate — it is to contract with hospitals for gate access, so the platform arrives as an invited partner rather than an intruder. This must be solved commercially before it is solved in software.
| Who | What happened |
|---|---|
| URAL EMS (2018) | Launched as "Bangladesh's first app-based ambulance service." Its Play Store listing now 404s and its domain no longer resolves. Dead. |
| SafeWheel | Started with on-demand rural ambulances — pivoted away to a medicine-delivery marketplace. |
| AmbuFast (Jan 2025) | Alive but shallow: ~2 App Store ratings, no update since April 2025. Funded by a retail-investor scheme, not institutional VC. |
| TraumaLink (NGO) | Not a competitor — a proof of concept. A volunteer SMS-dispatch network achieved ≤5-minute response in 88% of incidents across 3,119 patients. Distributed dispatch demonstrably works in Bangladeshi conditions. |
| Read this correctly. A weak field does not mean an easy market — it means nobody has proven the unit economics yet. Ambulance is a trust asset, not a revenue engine: it is a once-a-year emergency with poor retention. It earns the brand; diagnostics and tourism pay the bills. | |
Demand driver — road deaths, and why you must cite two numbers. Road Safety Foundation counted 7,359 killed in 2025; the official BRTA figure is ~5,490; the WHO estimates 31,578. The 4–6× gap between official and WHO figures is itself a well-documented fact. Cite both and name the discrepancy — using only one will get you challenged.
After a near-two-year freeze following August 2024, India resumed all visa categories for Bangladeshi nationals on 28 June 2026 — roughly three weeks before this dossier was written. All five IVACs moved to an automated time-slot system from 1 July 2026.
Any plan built on "India is closed and patients are stranded" is already obsolete. The rerouting to Thailand and China is real and it created the opening — but the corridor is reopening right now. The module must be multi-destination from day one, with India modelled as recovering rather than gone.
Sources: BSS (Bangladesh state news agency), Business Today, bdnews24, The Daily Star — June/July 2026.
| Destination | Reality | Cost level |
|---|---|---|
| 🇮🇳 India | ~50–60% of patients. Land border, no visa fee (Tk 1,500 processing). Reopened 28 June 2026. During the freeze, Kolkata hospitals lost up to 75% of volume — proof of how dependent they are on this flow. | Cheapest |
| 🇹🇭 Thailand | ~65,000 patients/yr. 60% of all Bangladeshi visa applications to Thailand are medical. MedPark Bangkok reports ~300% YoY growth. But ~80% of patients consider Thailand after an Indian visa failure and abandon it once they learn the cost. | 10–15× India |
| 🇨🇳 China (Kunming) | New and state-backed. ~600 patients treated by Aug 2025. One-day "green channel" visas. Bengali/English interpreters, halal food, dedicated wings. | ~¼ of Thailand |
| 🇲🇾 Malaysia | Record RM 10.4m revenue from Bangladeshi patients (2025). MHTC held Malaysia Healthcare Week in Dhaka (April 2026) and its CEO publicly said they want to partner with "local healthcare facilitators." | Mid |
| Best-sourced cost fact in the dossier (journalism, not facilitator marketing): liver transplant workup — India $1,000–2,000 vs Thailand $10,000–15,000. Cardiac stent — India ~$2,000 vs Thailand $5,000–20,000. Al Jazeera, Jan 2025. This 5–10× gap explains why patients cannot simply switch destinations. | ||
Kunming's binding constraint is not medicine or cost. It is language — and the price of solving it is explicit and public.
Bundle a vetted Bengali-speaking interpreter into a Kunming package and you have solved a concrete, priced, monetisable problem — not a slogan. This is the most actionable single finding in the dossier.
Rhythm Group (Bangladesh) × Manipal Hospitals (India) signed a medical-tourism deal in April 2026 — explicitly including a digital video-consultation platform so Bangladeshi patients can consult Indian specialists before travelling.
That is our tourism module, roughly three months ahead of us. It goes in the deck by name. A client will find it in thirty seconds; better they hear it from us, with a differentiation answer ready (multi-destination, published fees, interpreter bundle).
Facilitators typically earn 10–15%, up to 30% at the extreme; quotes are commonly padded 25–35% to absorb it. Critically, the fee is baked invisibly into the bundled price the patient is quoted — the patient never writes a separate cheque and generally does not know the commission exists. That opacity is exactly what a transparent platform disrupts.
⚠️ No verified Bangladesh-specific commission rate exists in any public source. This is the single most important number in the revenue model and it is not published anywhere. We model 10–15% and label it an assumption — never a fact. It must be confirmed by talking directly to facilitators and hospital international-patient departments.
This is the finding most likely to change your plan — and the one a competitor's deck will never tell you. The bundle is not the moat.
| Player | Bundle | Status & weakness |
|---|---|---|
| Sasthya Seba | Listing · Telemed · Lab · Ambulance · Medical tourism | Ships our exact feature list — on a €5,000 grant. Essentially unfunded, call-centre-driven, weak brand. Living proof that the bundle isn't the differentiator. |
| Shukhee (Grameen) | Listing · Telemed · Lab · Ambulance | Launched Nov 2024 with Grameenphone + Robi distribution. 2.8M installs but only ~49,000 consultations — a 1.7% activation rate. Telco-bundle enrolments are not customers. |
| Arogga | Pharmacy → Lab → Teleconsult | $5.7M raised — roughly 45% of all Bangladeshi health-tech funding ever. Has the transaction, the repeat customer and the logistics. The most dangerous competitor. |
| Praava Health | Own clinics + labs + telemed | $10.6M (2021) — best funded. Owns clinics, so owns the burn. Has been hunting a $15M Series B for ~5 years without closing it. |
| Doctorola | Doctor listing + appointments | The cautionary tale. Founded 2015, raised $250K. In 2026: 8 employees, zero press since 2016. It didn't blow up — it ossified. |
| Tonic (Grameenphone) | Telco health membership | ☠️ Dead. ~6M "members." Free service killed July 2021; IVR killed August 2021. |
| Rhythm × Manipal | Medical tourism + pre-travel video | 🔴 New — April 2026. Building our highest-margin module. |
Bundle into a free telco tier, report millions of "members," discover none will pay when the free tier ends. Tonic died of this. Shukhee is repeating it live.
Doctors won't pay for leads; patients won't pay booking fees; you become a call centre. Doctor listing is a feature, not a business.
Optimise for donor metrics ("lives impacted") and you build something that cannot survive without donors.
Own the clinics and you own the burn. Praava raised the most money in the sector and still can't close a Series B.
Bangladesh startup funding: $435M (2021) → $72M (2023) → ~$41M (2024), a six-year low. 2025's "$124M" headline is a mirage — $110M of it was a single late-stage M&A.
Bangladeshi health-tech has raised $12.6M in total, ever, across 15 funded companies. This business must reach profitability on revenue. Nothing may depend on raising a Dhaka Series A.
| Company | Where the money actually comes from |
|---|---|
| Practo (India) | ~35% SaaS sold to doctors (EMR/billing), teleconsult, corporate wellness. FY25 was its first full year of profitability. |
| Tata 1mg (India) | Pharmacy + own-lab diagnostics margin. ₹2,392 crore FY25. |
| Apollo 24/7 (India) | Works because it feeds an owned hospital chain. |
| Ping An Good Doctor (China) | Became profitable by selling to insurers and employers instead of consumers. 2025 net profit RMB 379.5m (+366%). |
| Halodoc (Indonesia) | Pharmacy + insurer/corporate deals. ~US$210M revenue FY2025. |
| Not one of them makes money from consumers booking doctors. Referral and booking are the acquisition layer — never the revenue layer. Any model that forgets this dies as Doctorola did. | |
This section is the most valuable in the dossier. Most competitors either do not know these rules or are quietly breaking them.
It is currently professional misconduct, not a crime, and it binds the doctor rather than the platform. But a platform that induces it can be blacklisted, and a Health Protection Act is being drafted that may criminalise the arrangement. Build so that a statutory ban would not kill the business.
BMDC Code §3 captures publicity by "anybody acting on his behalf" — our marketing of a doctor is caught by it.
⇒ Build a neutral directory card: permitted fields only, uniform styling, sorted by objective non-comparative criteria (availability, specialty, location, fee) — never by rating or paid rank.
| Area | The rule | What it forces |
|---|---|---|
| Telemedicine | Paid video consults are explicitly legal (BMDC Telemedicine Guidelines 2020, §5.8.3.1). But the platform must verify BMDC registration, display every doctor's registration number, provide a grievance channel, and record consent. | Verification is a legal duty, not a feature. |
| AI | §7.4 — AI/ML platforms may NOT counsel patients or prescribe. AI may assist and support only; the doctor must deliver the prescription. | No AI symptom-checker that diagnoses. Assistive only. |
| E-prescription | Legal, but bound by a drug matrix: List O (any mode) · List A (VIDEO ONLY) · List B (follow-up) · narcotics/psychotropics never. Drug names in CAPITALS. | The matrix must be hard-coded into the prescribing UI so the software prevents the offence. |
| Data residency | DGHS Telehealth Guideline §5(10): overseas cloud hosting is prohibited unless the provider has a datacentre inside Bangladesh. Personal Data Protection Act passed 10 April 2026. | Cannot default to a Singapore/US cloud region. In-country hosting; NDHIE/SeHR integration. |
| Payments | Holding patient money and settling it out to hospitals = Payment System Operator territory. Tk 5,00,000 licence, trust account, directors personally liable for shortfalls. | Never touch third-party funds. Use a licensed gateway's split-settlement / merchant-of-record model; invoice our fee separately. |
| Verification tooling | BMDC has a form-based lookup, no public API, no bulk registry. DGHS facility lists are stale PDF dumps. | Verification is a manual ops workflow with stored evidence and expiry tracking — not a scraper in the critical path. |
| ⚠️ Platform licensing | The DGHS National Telehealth Guideline states that independently operating telemedicine apps need licensing, with a price list approved by the authority. We could not confirm it is gazetted or that the licensing window is operational. Assume a licence is required; confirm with DGHS and counsel before launch. | |
The profitable model is illegal. The legal model is boring. That is precisely why nobody has built it properly.
Every constraint above is a barrier to entry for anyone who wants to do this the lazy way — and most incumbents are, right now, quietly non-compliant (ratings, reviews, fee-splitting, offshore hosting). A platform built legally from day one can say something no competitor can safely copy: every doctor's BMDC number is shown, every facility's DGHS licence is checked, and every fee we earn is disclosed on the screen where you pay it.
Compliance is not the tax on this business. It is the product.
Stated plainly, because a plan that hides its unknowns fails at the worst possible moment. Nothing below appears as fact anywhere in the proposal.
The widely-circulated "$70M market / $170M by 2027 / $600–750M TAM" figures trace only to an unverifiable Scribd-hosted PDF. Statista's numbers are paywalled and unread. Excluded from every document.
No source disaggregates BMDC or DGHS data to Dhaka. We did not invent one. If Dhaka TAM matters, it requires manual extraction from the DGHS registry or a direct data request.
The pain is described extensively in Bangladeshi media, but no study quantifies it. Any deck claiming "X% of tests are unnecessary" is making it up. This is a genuine research gap — and running our own survey would be a proprietary, defensible asset.
The single most important number in the tourism revenue model — and it does not exist publicly. No source states what an Apollo or Bumrungrad actually pays a Dhaka facilitator. Modelled at 10–15% and labelled an assumption.
Highly plausible and consistent with the documented domestic dalal (broker) culture — but no investigative report documents it for outbound medical tourism. Presented as a hypothesis to validate with patient interviews, never as a cited fact.
No public source confirms hospital-chain willingness to accept a third-party referral platform or what take rate they'd accept. Chains with their own apps (Square, Evercare, Ibn Sina) have little incentive. The entire referral thesis rests on this, and it is unvalidated. Twenty primary calls, before any code.
To be answered in writing, before a line of product code is written. Phase 0 of the build exists for exactly this.
Every material claim in the proposal, campaign and pitch deck traces to one of these.
| Domain | Sources used |
|---|---|
| Health expenditure & OOP | World Bank / WHO Global Health Expenditure Database (indicators SH.XPD.CHEX.GD.ZS, SH.XPD.CHEX.PC.CD, SH.XPD.OOPC.CH.ZS), 2023 · Bangladesh National Health Accounts (BNHA) / BIDS, presented July 2024 · The Financial Express & The Business Standard (budget), 2025–26 |
| Supply side | BMDC registration statistics, Nov 2024 (via Prothom Alo) · BBS Survey of Private Healthcare Institutions, 2021 (via Dhaka Tribune) · DGHS registration & licence-renewal data, 2025 (via The Daily Star) · WHO health labour market analysis |
| Price variance | New Age, "Pvt diagnostic facilities charge at will," 12 August 2022 · The Business Standard, "Refer a patient, get a hefty commission" |
| Ambulance / EMS | Hossain et al., Int'l Journal of Critical Care & Emergency Medicine 8:138, 2022 · TraumaLink, Global Health: Science and Practice, 2022 · BRTA registration data (via The Financial Express, 2021) · BSS (999 service data), 2025 · Prothom Alo (syndicate fares), 2025 · jagonews24 & TBS (syndicate deaths), 2025–26 · Road Safety Foundation & WHO road-death data, 2025–26 |
| Medical tourism | Ahsan H Mansur, Governor of Bangladesh Bank (via The Daily Star), 2025 · DCCI, 2024–25 · India Ministry of Tourism / Bureau of Immigration arrival data · ORF, "Medical Tourism as a Pillar of India-Bangladesh Relations," 2025 · Al Jazeera, January 2025 (cost comparison) · BSS, Business Today, bdnews24 (visa reopening), June 2026 · MHTC / PR Newswire, April 2026 · Dhaka Tribune & Prothom Alo (Kunming), 2025 · TBS (Rhythm × Manipal), April 2026 |
| Competition | Tracxn (health-tech Bangladesh) · Crunchbase / PitchBook · TechCrunch (Praava, Maya funding) · The Business Standard (Arogga, Shukhee) · Grameenphone official notices (Tonic discontinuation), 2021 · LightCastle Partners Startup Ecosystem report, 2026 · Startup Bangladesh Investment Report 2025 |
| Regulation | BMDC Telemedicine Guidelines, July 2020 (full PDF read) · BMDC Code of Professional Conduct, Etiquette and Ethics (full text extracted) · DGHS National Telehealth Guideline (Feb/May 2026 documents, full PDF read) · Medical Practice and Private Clinics and Laboratories (Regulation) Ordinance, 1982 · Personal Data Protection Act (passed 10 April 2026) · Cyber Security Ordinance 2025 · Bangladesh Payment and Settlement Systems Regulations / draft PSO Regulation 2025 · DGHS 8-point directive, 11 September 2025 |
Research conducted July 2026 across five parallel streams. Where two official sources conflict (health expenditure OOP share; facility counts; road deaths), both are presented and the conflict named rather than resolved by averaging. Where a figure could not be traced to a credible primary source, it has been excluded — see §09.
