ola.tech — Agentic Second-Hand Phone Global Trade

I. The Thesis

Raymond Sze recently took over ola.tech (legal entity: FonFair Technology Limited), a Hong Kong–based refurbished phone e-commerce site with a few years of operations and 20,000+ claimed customers.¹ The site sources refurbished iPhones and Android devices from Japan, USA, and HK, selling primarily to HK consumers at 50–90% off retail.² SimilarWeb confirms 14.2K monthly visits (Jan 2026) with 47.5% from search, 40.5% direct, and a global traffic split: HK 38%, India 24%, US 21%, Indonesia 9%, Pakistan 6%.²⁴

The asset came with something more valuable than the storefront: active WhatsApp inbound leads from around the world — buyers looking to purchase second-hand phones. The traffic geography confirms global reach: India, Indonesia, and Pakistan alone represent 38% of visitors, all high-demand markets for refurbished phones. Raymond has since moved to Shenzhen, placing him minutes from Huaqiangbei/Huaqiangnan — the world’s largest used phone trading hub, responsible for over 70% of globally refurbished iPhones.³

The question: Can Raymond (Shenzhen sourcing, prior SZ phone trader relationships) and Eric (HK, agentic infrastructure) use AI agents to handle WhatsApp-based global sales at scale — effectively running a lean cross-border phone trading operation with near-zero headcount?

Dog-food signal: Moderate — stronger than first assumed

Raymond has direct pain: He previously ran a Filipino retail/e-commerce business in HK, sourcing and selling second-hand phones to HK Filipinos. He’s dealt firsthand with SZ phone traders and knows the trust problem — traders swap components, misrepresent grades, maximize profit at your expense since they don’t expect repeat business.
The agentic infrastructure IS something Eric is building (Donna, Sourcy WA bot). The WhatsApp agent for ola.tech is a direct application of existing work.
Verdict: Raymond has the sourcing pain, Eric has the tool-building trajectory. Not a pure dog-food play on the product side, but the operational pain is real and firsthand.

Monthly Traffic

14.2K

SimilarWeb Jan 2026

Bounce Rate

40%

Decent for e-commerce

Pages/Visit

1.77

Low — browsing not deep

Avg Duration

0:22

Quick — price-check traffic

Traffic quality flag: 22-second average visit + 1.77 pages = price-check traffic, not deep shopping.

Top keywords are branded (“ola tech”, “ola hk”) and generic (“iphone”, “13 pro white”) — people searching for cheap phones, landing, checking price, leaving.
The real conversion channel is likely WhatsApp, not the website checkout. Site serves as a catalogue; WhatsApp is where trust is built and orders close.
This actually supports the agentic thesis: the website generates awareness, WhatsApp closes the deal.

Global TAM (2025)

US$65–74B

315M+ units/yr

India+ID+PK Traffic

38%

High-demand markets

B2C Gross Margin

48–60%

ola.tech current model

B2B Wholesale

5–15%

Thin but volume-driven

II. Founder Context

Dimension	Raymond Sze	Eric San
Location	Shenzhen (recently relocated)	Hong Kong
Domain expertise	HAS EXPERIENCE Previously ran Filipino retail/e-commerce in HK — sourced and sold second-hand phones to HK Filipinos. Has existing relationships with SZ phone traders. Knows the trust problem firsthand (component swaps, misrepresented grades). Co-founded BornTea with Leo Wong.	None in phone trading. Builder — Donna, OpenClaw, Sourcy WA bot, PCRM.
Sourcing access	STRONG Lives near Huaqiangbei. Has existing SZ trader relationships. Can physically visit wholesale markets (Tongtiandi, Mingtong, Yuanwang). Knows the scams — key tacit knowledge.	HK-based, can handle logistics to/from HK airport hub.
Tech capability	Tech enthusiast but not builder.	STRONG Building WhatsApp agents, already has WA bot architecture from Sourcy.
Capital	Unknown. Phone trading is working-capital intensive.	Fractional. Not primary financial backer.
Network	GenieFriends. Leo Wong (imBee, BornTea co-founder) — see below.	Broad. Sourcy (supply chain), Kimberly (OEM/import certs), Shannon (Amazon import certs, China-Cambodia).

Correction from v1: Raymond HAS phone-trading experience.

He sourced and sold second-hand phones to HK’s Filipino community via e-commerce. This is directly relevant experience.
He knows the SZ trader dynamics: “they will find every opportunity to max profits at your expense — components may be different from advertised — since they don’t expect to see you again.”
This changes the risk profile significantly. The experience gap is narrower than initially assessed. The key gap is now scale (small HK Filipino e-commerce → global cross-border trade) and working capital.

Leo Wong — potential third leg (status: unclear if available for this project)

imBee (Co-founder, Jan 2020–present): AI-Powered Chat Management Platform — unified communication SaaS for businesses. DIRECTLY RELEVANT to WhatsApp commerce infrastructure.
BornTea (Co-founder with Raymond, 2017–present): Chinese tea leaf retail brand.
Sensbeat (Co-founder/CEO, 2013–2016): Music app. Team of 18. US$1M seed raised during college. 200K+ downloads. 5 international awards. HK/Beijing/SV.
HKUST BBA (Marketing & IS). 5,150 LinkedIn followers. Serial entrepreneur.
Why it matters: Leo literally builds chat SaaS for businesses. If he were involved, the WhatsApp agent infrastructure would be a trivial build, not a project. But his availability and interest are unconfirmed — imBee is his primary commitment.

III. Market Sizing

Global TAM

Source	2025 Value	2030 Projection	CAGR
Mordor Intelligence⁴	US$65.2B	US$91.1B	6.93%
Deep Market Insights⁵	US$73.9B	US$120.2B	10.2%
Research & Markets⁶	US$35.8B	US$53.3B	8.24%
IDC (volume)⁷	~415M units	—	~6%

Wide variance between sources (US$36B to US$74B) reflects different scope definitions — some include feature phones, some only smartphones, some include accessories. CONSENSUS: US$50–75B, 300–415M units, growing 6–10% annually.

Regional TAM — Where the Demand Lives

Asia-Pacific

~45% of global

Africa + Middle East

US$6.1B → $12.4B

Europe

~22%

North America

~15%

Africa is the fastest-growing demand market for used phones, driven by mobile-first economies, youthful demographics, and affordability needs.⁸ Nigeria, Kenya, Ghana, Egypt, and South Africa are the top markets. Buyers prefer Grade B/C devices, dual-SIM, unlocked models — exactly what Huaqiangbei supplies.⁹

Local TAM — Hong Kong (ola.tech’s Current Base, Not Its Ceiling)

HK’s physical second-hand phone market has collapsed. Sin Tat Plaza in Mong Kok — once home to 300+ phone stalls — is now “a dead town” with most stalls closed and a poor reputation for authenticity.¹⁰ But SimilarWeb reveals ola.tech is already global: only 38% of traffic is HK. India (24%), US (21%), Indonesia (9%), and Pakistan (6%) together outweigh HK.²⁴ The site has organically attracted exactly the markets that need refurbished phones most.

Addressable TAM — What Raymond + Eric Can Realistically Reach

The thesis is not to compete with Back Market or ATRenew at scale. It’s to use SZ sourcing + AI agents to service WhatsApp-based inbound demand from buyers in Africa, Middle East, and Southeast Asia — markets where trust is built on WhatsApp conversations, not marketplace listings.

Realistic Year 1 addressable: US$500K–2M in GMV

Assume 50–200 orders/month at US$200–800 average order value (mix of B2C singles and small B2B batches).
At 15–30% gross margin (blended B2C + B2B), that’s US$75K–600K gross profit.
This is a lifestyle business at the low end, a serious SME at the high end.

IV. Competitive Landscape

4a. The Giants — Marketplace Players

Company	Model	Revenue / Scale	Playbook	Why It Doesn’t Apply
Back Market¹¹ $5.7B valuation	Marketplace. 10% commission + services. 2,700 vetted refurbishers.	US$415M rev (2024). US$2.8B GMV. 17M customers. 18 markets.	Massive paid acquisition + brand trust building over 10+ years. $1B raised.	INAPPLICABLE Took $1B and 10 years. Marketplace model requires massive two-sided network effects.
Swappie¹² Finland	Buy-refurbish-sell (vertically integrated). iPhone-only.	€249M rev (2024). 617K units. Still unprofitable (−€24M net loss).	Own the refurbishment. Control quality. European EIB loan (€18.4M).	STILL UNPROFITABLE at €249M revenue. Needs massive scale for unit economics to work. Capital-intensive model.
ATRenew¹³ China (NYSE: RERE)	Full stack: C2B trade-in + B2B auction + B2C retail.	RMB 5.1B/quarter (~US$720M). 10.9M units Q3 2025. Profitable.	Dominant China position. JD.com partnership for trade-ins. Auction platform (PJT) for B2B distribution.	CHINA ONLY — doesn’t serve global buyers via WhatsApp. Different beast entirely. But proves the SZ-sourcing economics work at scale.
Reebelo¹⁴ Singapore	Marketplace for certified refurbished electronics. APAC-focused.	US$52M raised (Series A). 70K+ Trustpilot reviews (4.7/5).	APAC-first, certified sellers, Antler-backed.	CLOSEST COMP in region. But they’re a marketplace — ola.tech is a direct seller. Different margins, different game.
Alibaba wholesale¹⁵	B2B marketplace. SZ wholesalers list directly.	2,766+ mobile phone listings. Trade Assurance protection.	Volume, trust (escrow), supplier verification.	DIRECT COMPETITOR to B2B wholesale angle. Alibaba is where African/ME buyers already go. Hard to out-Alibaba Alibaba.

4b. Playbook Dissection — What Actually Works

Company	What Worked	Lesson for ola.tech
ATRenew	Vertically integrated: collects phones (C2B via JD.com), grades them (proprietary AI grading), auctions to B2B buyers (PJT marketplace), sells remainder B2C (Paipai). The auction platform is where the real margin lives.	KEY INSIGHT The B2B auction/distribution layer is the profitable part. ATRenew’s PJT marketplace is essentially what ola.tech’s WhatsApp channel could become — connecting SZ supply with global B2B demand.
Back Market	Didn’t touch inventory. Pure marketplace with trust layer (quality scoring, warranty, customer protection). The 10% take rate is pure margin.	Trust is everything. ola.tech’s 90-day warranty and 25-point inspection are table stakes. Need to build trust via WhatsApp rapport before buyers commit to cross-border purchases.
AUCNET (Japan)¹⁶	Weekly online auctions of 30,000+ used phones to 1,700+ companies worldwide. B2B only. Professional grading with Blancco data erasure certification.	MODEL TO STUDY This is essentially what a scaled version of ola.tech’s B2B channel looks like. Weekly auctions, professional grading, global distribution.

4c. Failed Examples — The Startup Graveyard

Company	What Happened	Lesson
Refurbished.nl¹⁷ Netherlands, 2024	Declared bankrupt May 2024. Founded 2016 as refurbished electronics marketplace. Failed due to: intense competition, exorbitant marketing costs, surge in customer complaints about undelivered products, unpaid returns, rent arrears.	MARKETING COST DEATH — CAC in marketplaces is a race to the bottom. ola.tech’s WhatsApp inbound avoids this trap (leads come to you), but only if the inbound volume is real.
Leapp¹⁸ Netherlands, 2018	Refurbished Apple retailer. 24 stores across NL/BE/DE. Bankrupt from operating losses after rapid physical expansion. Rescued by PE, scaled back to 14 stores.	PHYSICAL EXPANSION DEATH — Not relevant to ola.tech (online-only). But validates that refurbished phone retail has thin margins that don’t survive overhead.
Swappie¹² Finland, ongoing	Still operating but burning cash. €249M revenue, −€24M net loss in 2024. Raised $124M+ plus €18.4M EIB loan. 617K units/year.	SCALE DOESN’T FIX MARGINS — Even at 617K units, Swappie loses money. Vertically integrated refurbishment is capital-intensive. ola.tech should NOT try to own refurbishment — source pre-graded from SZ wholesalers.

The pattern: marketing costs and thin margins kill refurbished phone companies.

Refurbished.nl died from CAC. Leapp died from overhead. Swappie is bleeding at €249M revenue.
ola.tech’s advantage: inbound WhatsApp leads = near-zero CAC. If this signal is real and sustainable, it’s the single most important asset.
The question becomes: where do the WhatsApp leads come from, and will they keep coming?

V. Unit Economics

Revenue Side — ola.tech B2C (Current Model)

Metric	ola.tech (estimated)	Benchmark	Source
Average selling price (ASP)	HK$2,500–4,000 (~US$320–515)	US$176 avg used phone (IDC global)	⁷
SZ wholesale cost (Grade A)	US$150–300 (iPhone 13–15)	Varies by model/grade	³
Gross margin (B2C)	48–60%	20.3% (Swappie, includes refurbishment)	¹²
Gross margin (B2B wholesale)	5–15%	2–25% range (industry)	¹⁹
Shipping per unit (SZ→HK)	~US$3–5	SF Express/courier	²⁰
Shipping per unit (global)	~US$8–25	DHL/SF intl express	²⁰
Return rate	~5–10% (industry avg)	8–12% for refurbished electronics	¹⁹

Cost Side — COGS Breakdown Per Unit

Cost Component	Per-Unit Cost	Assumption
Phone procurement (SZ wholesale)	US$150–300	iPhone 13/14 Grade A from Huaqiangnan
QC / grading	US$2–5	Basic inspection. No refurbishment — buy pre-graded.
Shipping to customer	US$8–25	Varies: HK $3, Asia $8, Africa/ME $15–25
Packaging	US$1–2	Box + cable included
WhatsApp API (sales agent)	US$0	FREE — customer-initiated service messages cost nothing.²¹
AI inference (per conversation)	US$0.01–0.05	GPT-4o-mini / DeepSeek for sales conversations
Payment processing	~2.5–3.5%	Stripe / PayPal for intl. payments
Warranty claims (~7%)	US$15–35 amortized	90-day warranty. Replacement or partial refund.

The death cost: Working capital, not COGS.

Phone trading is a working capital business. You buy inventory before you sell it.
At 100 phones/month at US$200 avg procurement: US$20,000/month tied up in inventory.
At 500 phones/month: US$100,000/month. Cash cycle is 2–6 weeks (buy → ship → receive payment).
This is the #1 killer of small phone traders. Not margins, not marketing — cash flow.

Scenario Analysis — Monthly P&L

Pessimistic

30 orders · US$1.5K profit

Realistic

100 orders · US$6K profit

Optimistic

300 orders · US$22K profit

Assumes blended 20% net margin after all costs. B2C singles at higher margin, B2B batches at lower margin. Pessimistic scenario barely covers Raymond’s living costs in SZ. Realistic is a solid SME income for two people.

Break-Even

Assuming US$3,000/month fixed costs (SZ rent, tools, hosting), break-even requires approximately 50 orders/month at US$60 avg net profit per order. This is achievable within 3–6 months if WhatsApp inbound is real.

VI. The Agentic Angle — Why This Is Different

The conventional refurbished phone business requires: salespeople (to answer inquiries), inventory managers (to track stock), logistics coordinators (to arrange shipping), and customer service reps (to handle returns). A 5–10 person team for a small operation.

The agentic thesis: replace 80% of that headcount with AI agents on WhatsApp.

What the Agent Does

Function	Human Today	Agent Tomorrow	Cost Delta
Answer inbound inquiries	1–2 salespeople	WhatsApp AI agent (24/7, multilingual)	US$3K/mo → US$0.05/convo
Quote pricing	Manual lookup + negotiation	Agent queries live inventory DB, generates quote	Minutes → seconds
Order processing	Manual data entry	Agent creates order, triggers payment link	US$0 (automated)
Inventory sync	Spreadsheet or basic ERP	Agent reads SZ supplier APIs or shared inventory sheet	Real-time vs. daily
Customer follow-up	Manual reminders	Agent sends shipping updates, warranty reminders	US$0 (service messages free)

WhatsApp pricing is the enabler.

Customer-initiated service messages: FREE (since Nov 2024).²¹
Since ola.tech’s model is inbound (customers reach out first), virtually all conversations are free.
Only outbound marketing messages cost money (US$0.025–0.14 per message) — and you rarely need those for inbound-driven business.
Compare: a human salesperson handling 20 WhatsApp conversations/day costs US$2,000–3,000/month. An AI agent handles 200/day at <US$10/month in inference costs.

Existing Tools in the Space

Several startups are building WhatsApp commerce infrastructure: Zotok.ai (order-to-cash automation, ERP integration), QuoteSenderBot (automated quoting from CSV), Galo (B2B food supply chain on WhatsApp), and ZOLO (food supplier automation).²² None are specifically built for used phone trading, but the infrastructure exists. Eric could either build custom (using his existing Sourcy WA bot architecture) or adapt an existing tool.

Eric’s unfair advantage: He’s already building this.

The Sourcy WhatsApp bot is the same architecture — answer product inquiries, handle trust-building, process orders.
Eugene Clarance (Sourcy) already shared: “most leads test trust (price/MOQ/delivery), not asking for recommendations.” The same pattern applies to phone trading.
The agent doesn’t need to be fancy. It needs to: quote a price, show product photos, answer “is this legit?”, and process payment.

VII. GTM Assessment

Phase 0 — Validate (Week 1–4)

Before building any AI agent, Raymond should manually handle 20–30 WhatsApp conversations. This does three things:

Confirms the inbound leads are real buyers, not spam/bots.
Maps the buyer profile: where are they? What do they want? What price range? B2C or B2B?
Builds the “sales script” that the AI agent will later replicate.

The single most important question: Are the WhatsApp leads real?

If yes → this business has legs. Inbound leads with near-zero CAC is the holy grail.
If no (spam, bots, window-shoppers) → the entire thesis collapses. Don’t build anything until this is validated.
Raymond should track: lead source (country code from WA number), inquiry type (B2C vs B2B), average order intent, conversion rate to payment.

Phase 1 — Manual B2C + First B2B (Month 2–4)

Continue ola.tech B2C for HK customers (existing model, proven).
Manually fulfill first B2B orders from WhatsApp leads (Africa, ME, SE Asia).
Raymond visits Huaqiangnan wholesalers, builds 3–5 supplier relationships for different grades/brands.
Document every step: what questions buyers ask, what photos they want, how they pay, what shipping method they prefer.
Eric begins building WA agent based on these real conversation patterns.

Phase 2 — Agent Deployment (Month 4–6)

Deploy WhatsApp AI agent for initial inquiry handling (quoting, inventory check, FAQ).
Human-in-the-loop for: payment processing, complex negotiations, dispute resolution.
Target: agent handles 70% of conversations autonomously, escalates 30%.
Begin tracking: response time, conversion rate (agent vs. human), customer satisfaction.

Phase 3 — Scale B2B (Month 6–12)

If B2B wholesale works → focus on repeat buyers (African distributors, ME retailers).
Weekly inventory list sent via WA broadcast (automated from SZ supplier inventory).
Consider AUCNET-style mini-auctions for B2B buyers (WhatsApp group auction format).
Working capital: need US$50K–100K credit line or partner capital to scale B2B.

Government Grants

BUD Fund is eligible for HK-registered companies expanding to ASEAN and mainland China.²³ Up to HK$7M cumulative at 50% matching. Could cover technology adoption (WA agent development), market expansion (Africa, ME), and branding. ola.tech (FonFair Technology Limited) should be eligible if it has substantive HK operations.

TVP (Technology Voucher Programme) closed to new applications after Dec 2024.²³

VIII. Red Team — Why This Might Not Work

Why It Could Work

Existing asset: ola.tech has brand, domain, SEO, 20K+ customers, and live WhatsApp leads. Not starting from zero.
SZ proximity: Raymond is minutes from the world’s cheapest phone sourcing. Direct access = no middleman markup.
AI agent + WhatsApp = near-zero sales cost. Service messages free. Inference <US$0.05/conversation. Scales without headcount.
Inbound demand: The hardest part of any business (finding customers) may already be solved. If leads are real, this is the holy grail.
Low barrier to test: Can validate with US$5K and 30 days. No hardware, no complex build.
Massive underserved market: Africa + ME phone demand is growing 10%/yr and is poorly served by online marketplaces.
Raymond has done this before. He sourced and sold phones to HK Filipinos. He knows SZ trader dynamics, scam patterns, and grading. Not starting from zero.
Leo Wong (imBee): Raymond’s BornTea co-founder runs a chat management SaaS. If he even advises on WhatsApp infrastructure, the agentic build gets dramatically easier.

Why It Might Fail

Limited scale experience: Raymond has traded phones for HK Filipino market, but not at global cross-border scale. Grading at volume, multi-currency payment, and intl. logistics are new.
Working capital trap: Phone trading eats cash. 100 phones/month = US$20K tied up. Without capital, can’t scale.
Fraud risk: Cross-border phone trading is rife with scams — both from buyers and suppliers. IMEI blacklisting, stolen phones, non-payment.
Alibaba competes directly: B2B wholesale buyers can already find SZ suppliers on Alibaba with Trade Assurance protection. Why use ola.tech?
Trust is earned, not automated. African/ME buyers build trust through WhatsApp rapport — over weeks, not minutes. An AI agent may feel impersonal for high-value B2B orders.
Returns destroy margins. Cross-border returns for refurbished phones are logistically nightmarish and expensive. A 10% return rate at 5% margin = negative economics.
WhatsApp leads may be ephemeral. If the previous operator generated them through marketing/SEO that isn’t maintained, they’ll dry up.

Steel-Man: The Strongest Case Against

The strongest version of the bear case: Alibaba Trade Assurance already solves the trust problem for B2B phone wholesale. A buyer in Nigeria can find a verified SZ supplier, get escrow protection, and buy at wholesale prices — all without needing ola.tech as an intermediary. ola.tech would be a more expensive middleman with less protection. The only counter: ola.tech’s WhatsApp agent could provide a more personalized, responsive experience than Alibaba’s clunky platform — and for smaller buyers (5–50 units), the personal touch matters more than Trade Assurance.

Steel-Man: The Strongest Case For

The strongest bull case: The phone trading business is already validated. ola.tech has years of operations and thousands of customers. Raymond isn’t building something new — he’s revitalizing an existing business with better sourcing (SZ direct) and better operations (AI agents). The risk isn’t “will anyone buy refurbished phones?” (obviously yes, it’s a $65B market). The risk is purely execution. And the cost to test is near-zero.

IX. Verdict

CONDITIONAL YES — worth experimenting, but sequence matters.

ola.tech is not a startup idea — it’s an existing business with real assets: 14.2K monthly visits, global traffic (India/Indonesia/Pakistan = 38%), inbound WhatsApp leads, and a founder who has actually sourced and sold phones from SZ before. The question isn’t “is the market real?” — it demonstrably is (US$65B+ globally, growing 7–10%/year). The question is: can you scale from a small HK Filipino e-commerce operation to global cross-border trade, using AI agents as the operational backbone?

The answer depends on exactly one thing: are the WhatsApp inbound leads real, paying buyers?

If yes — this is a genuine cashflow-positive SME opportunity within 3–6 months. Raymond’s existing SZ trader relationships provide the sourcing advantage. Eric’s WA agent architecture provides operational leverage. The combination of near-zero CAC (organic inbound from 14.2K monthly visits) + near-zero ops cost (AI) + best-in-world sourcing (Huaqiangbei) + founder who knows the scam patterns is structurally attractive. Realistic Year 1: US$500K–2M GMV, US$75K–300K gross profit for two people.

If no — the thesis collapses. Without organic inbound converting to WhatsApp orders, you’re back to competing on marketing spend against Alibaba, Back Market, and Reebelo — a losing game for a two-person team.

The minimum viable test: Raymond handles 20 WhatsApp conversations manually over 2 weeks. Track: buyer country, order intent (B2C/B2B), price sensitivity, conversion to payment. Cost: US$0 (just time). If 5+ convert to real orders, proceed to Phase 1.

What would upgrade to strong yes: (1) WhatsApp leads convert at 10%+ to paid orders. (2) Leo advises on WA infrastructure (his imBee is literally chat SaaS). (3) First B2B repeat buyer established (proves the flywheel).

What would change the answer to a no: WhatsApp leads turn out to be mostly spam or window-shoppers, OR working capital requirements exceed what Raymond can fund, OR the 22-second average visit duration reflects a site with no purchase intent.

References

[1] ola.tech — Main site. Claims 20,000+ customers, 100,000+ renewed devices. Entity: FonFair Technology Limited.

[2] ola.tech Help Center — FAQ: devices sourced from Japan, USA, HK. 90-day warranty. 25-point inspection. Battery 80–100%. WhatsApp + Messenger support.

[3] The7 Electronics — Huaqiangbei Used Phone Market Guide — Huaqiangnan is the actual used phone hub. Over 70% of globally refurbished iPhones sourced here.

[4] Mordor Intelligence — US$65.2B (2025) → US$91.1B (2030), 6.93% CAGR. 315M→430M units.

[5] Deep Market Insights — US$73.9B (2025) → US$120.2B (2030), 10.2% CAGR. APAC largest and fastest-growing.

[6] Research & Markets — US$35.8B (2025) → US$53.3B (2030), 8.24% CAGR. Narrower scope (smartphones only).

[7] IDC — ~415M used smartphones shipped in 2026, market value US$99.9B. 195M units in 2023 (+6.4% YoY).

[8] Custom Market Insights — ME+Africa: US$6.11B (2022) → US$12.39B (2030), 10.01% CAGR.

[9] Recirq Global — African buyers prefer Grade B/C, dual-SIM, unlocked. Nigeria largest market. iPhone + Samsung = 64% share.

[10] HK Free Tours — Sin Tat Plaza — Once 300+ stalls, now mostly closed. Poor reputation for authenticity. “Dead town.”

[11] Sacra — Back Market — US$415M rev (2024), US$2.8B GMV, US$5.7B valuation. $1B raised. 17M customers. 10% commission + 4.8% services take.

[12] AIM Group — Swappie 2024 — €249M rev, 617K units, -€24M net loss. 20.3% gross margin. €18.4M EIB loan.

[13] ATRenew Q2 2025 — RMB 5.0B/quarter (~US$697M). 32.2% YoY growth. Income from operations RMB 91.1M. NYSE: RERE.

[14] Reebelo Series A — US$52M raised. Cathay Innovation led. APAC-focused marketplace. 70K+ Trustpilot reviews (4.7/5).

[15] Alibaba — Mobile Phones in Africa — 2,766+ listings. SZ wholesalers. Trade Assurance. Direct B2B competition.

[16] AUCNET Global Auction — Weekly auctions of 30,000+ used phones. 1,700+ companies worldwide. Blancco-certified data erasure.

[17] Silicon Canals — Refurbished.nl Bankruptcy — Bankrupt May 2024. Intense competition + exorbitant marketing costs + customer complaints.

[18] Silicon Canals — Leapp — Bankrupt 2018. 24 stores NL/BE/DE. Rescued by PE. Scaled back to 14 stores.

[19] TG Wireless — Wholesale Phone Margins 2026 — Flippers 2–5%, value-added 5–12%, full refurb up to 25%. New flagships 8–12%.

[20] SF Express — International Rates — SZ→HK and international rates. Multiple service tiers.

[21] Flowcall — WhatsApp API Pricing 2026 — Service messages FREE (customer-initiated). Marketing US$0.025–0.14/msg. Per-message billing since July 2025.

[22] Zotok.ai — WhatsApp order-to-cash automation. ERP integration. 5–15% sales acceleration claimed.

[23] HKPC — TVP & BUD — TVP closed Dec 2024. BUD active: HK$7M cumulative, 50% matching, covers ASEAN + FTA markets.

[24] SimilarWeb — ola.tech (Jan 2026) — 14.2K monthly visits. Bounce 39.97%. Pages/visit 1.77. Avg duration 0:22. HK 38%, India 24%, US 21%, Indonesia 9%, Pakistan 6%. Search 47.5%, Direct 40.5%.