A full Nigerian Pidgin translation of this report is not available yet — showing the English edition below.
Global Ecommerce Data Sources and Methodology
This page is the methodology and data-sources overview for Globegraphy research. It documents how we grade global ecommerce data sources, how the ecommerce research methodology works, and how to read figures on Atlas, region, and research pages. The audit covers 217 cited sources (August 2026): Grade A 81 (37%), Grade B 116 (53%), Grade C 17 (8%), plus 4 unclassified domains handled individually. Use it to trace every key number back to a graded citation.
Related reading: Europe Ecommerce Atlas · Middle East Ecommerce Atlas · Top Ecommerce Platforms in Europe & North America · Europe research findings.
Data Sources Overview
Globegraphy cites only sources found in the underlying worldwide ecommerce landscape report. Domains are graded A/B/C by authority (official statistics and primary research vs industry media vs aggregators). The grade table and domain inventory are in §2 Domain Quality Classification below; the executive summary of counts and share is in §1.
| Item | Content |
|---|---|
| Audited | 全球电商格局研究报告.md (564 lines, 217 source URLs) |
| Date | 2026-08 (completed the same day) |
| Scope | Quality grading of all 217 source URLs; secondary verification of key data points; inventory of errors and high-risk sources |
| Method | ① Programmatically extract all URLs and grade them by domain; ② Independently verify key data points via web_search (3 parallel verification agents + human spot-checks); ③ Correct confirmed issues |
Methodology
Verification combines programmatic URL extraction, independent web checks by three verification agents, and lead-agent human spot-checks. Coverage and pass/fail detail are in §3 Data Point Verification. Prefer Grade A sources for decision-critical figures; treat Grade C Chinese aggregators as restatements that need a second check (see §5 Recommendations).
Data Vintage and Updates
- Audit date: August 2026 (same-day completion for the grading pass).
- Report vintage: figures in the audited landscape report are labelled with source and year; conflicting definitions are annotated rather than collapsed.
- Corrections already applied: truncated URLs, proxy links, and high-risk annotations are listed in §4 Corrections Applied.
Limitations and Caveats
- Regional market-size figures use multiple definitions (B2C vs retail vs GMV; with/without services). Keep definition notes when citing externally.
- Two figures could not be independently corroborated in this audit (Mexico 84% penetration; China CR5 78.3%) and remain annotated — see §1 and §4.
- Grade C / aggregator restatements have known transcription risks (including a prior 10× ECDB error on AMZ123). Trace back to the primary report before external use.
How to Use This Research
- Start from an Atlas or research page for the market question, then use this audit to check source grade and vintage for any figure you cite.
- For external decisions, prefer Grade A citations (eMarketer, ECDB, official statistics, company filings) listed in §2.1.
- When a figure carries a caveat in §4, carry that caveat into any client or investment use.
- Pair this page with the Europe Atlas, Middle East Atlas, and regional research notes for context — this page is methodology, not a market overview.
1. Executive Verdict
Overall quality: Good (B+). Of the report’s 217 source citations:
| Quality grade | Meaning | URL count | Share |
|---|---|---|---|
| Grade A — primary authoritative sources | Official statistics (NBS/METI/statistics bureaus/Ministry of Commerce), leading research houses (eMarketer/ECDB/Statista/Forrester/Bain/Juniper/GSMA), first-party company disclosures (SEC filings, official newsrooms), major wire services (Reuters/Guardian/Politico/Yonhap), and authoritative ecommerce data houses (DC360/Marketplace Pulse/PYMNTS) | 81 | 37% |
| Grade B — solid industry media / secondary sources | Industry media (ecommercenews/Retail Gazette/Cross-border Magazine), corporate newswire distributors (GlobeNewswire/Business Wire — quality depends on the original publisher, often Research and Markets reports), and financial media restating first-party data | 116 | 53% |
| Grade C — secondary Chinese aggregators / treat with caution | AMZ123, Chwang, DNY1 and similar Chinese cross-border trade media (known 10× transcription errors); Renren Doc (user-uploaded documents); SEO content sites | 17 | 8% |
| Unclassified | 4 domains (handled individually; see §4) | 4 | 2% |
Confirmed issues that have been corrected: 1. ✅ 11 truncated URLs (incomplete links ending in "...") replaced with full URLs. 2. ✅ 1 campus-proxy link (Euromonitor via HKUST ezproxy) replaced with a public URL. 3. ✅ 1 unknown-source domain (oudai1.com) replaced with a verifiable source. 4. ⚠️ Several figures from lower-quality sources have been annotated (China CR5, Mexico 84% penetration, mobile 73%/72%, social commerce $1.2T, India quick-commerce forecasts, etc.); see §4.
Key data points that passed verification during the audit (spot-checks): 2025 global $6.42T/+6.8%/20.5% (eMarketer); 2028 $7.89T/22.5% (eMarketer); TikTok Shop global $64B/+94%; ECDB 2025 company GMV ranking (Alibaba/PDD/Amazon); Chinese platforms ~50% of Southeast Asia GMV; Mexico 2024 MX$789.7B/+20%; US 2025 ecommerce $1.5T+/Q4 25%; Euromonitor article public link. See Section 3 for the verification detail.
Verification coverage:
| Channel | Items covered | Result |
|---|---|---|
| Verification agent B (Europe + Asia-Pacific) | 22 | 21 OK / 0 ISSUE / 1 UNVERIFIED (China CR5) |
| Verification agent C (LatAm + Middle East & Africa + trends) | 24 | 23 OK / 1 OK-with-caveat (Temu share is a cross-border metric) / 0 ISSUE |
| Verification agent A (global + North America) | 17 | 17 OK (lead agent human item-by-item review) |
| Lead-agent human spot-checks (cross-region) | ~24 | 21 OK / 2 UNVERIFIED (Mexico 84%, China CR5) / 1 conflict corrected (US online shoppers) / several definitional annotations |
In total, 63+ data points were verified. No material numeric errors were found (0 ISSUE); 2 figures could not be independently corroborated (both annotated), 1 definition needed clarification (Temu cross-border share, already corrected), and several lower-quality sources were replaced or annotated.
2. Domain Quality Classification
2.1 Grade A: primary authoritative sources (81 citations, 37%)
| Domain | Count | Notes |
|---|---|---|
| www.emarketer.com | 6 | eMarketer/Insider Intelligence — authoritative global ecommerce forecasts |
| ecdb.com | 5 | eCommerceDB — authoritative country/company/platform ecommerce data |
| www.reuters.com | 3 | Reuters |
| news.mercadolibre.com | 3 | Mercado Libre official newsroom |
| www.digitalcommerce360.com | 7 | Authoritative US ecommerce data (Census-based estimates) |
| www.marketplacepulse.com | 4 | Authoritative platform GMV/seller data |
| www.pymnts.com | 2 | Authoritative payments-industry media |
| www.globenewswire.com | 10 | Corporate newswires (mostly Research and Markets report releases; content is authoritative) |
| www.businesswire.com | 2 | Corporate newswire distribution |
| www.sec.gov | 2 | US SEC official filings |
| www.meti.go.jp | 1 | Japan Ministry of Economy, Trade and Industry |
| dzswgf.mofcom.gov.cn / cm.mofcom.gov.cn | 2 | China Ministry of Commerce |
| paper.people.com.cn | 1 | People’s Daily |
| en.yna.co.kr | 1 | Yonhap News Agency |
| en.baochinhphu.vn | 1 | Vietnam government website |
| www.nzpost.co.nz | 1 | NZ Post official |
| www.ibef.org | 1 | India Brand Equity Foundation |
| www.forrester.com / www.bain.com / www.juniperresearch.com / nielseniq.com / www.globaldata.com / www.ystats.com / www.mirakl.com / www.worldpay.com | 1–2 each | Leading research/consulting firms |
| corporate.zalando.com / en.media.allegro.pl / ir.coveo.com / www.shopify.com / www.aboutamazon.eu / blog.google / www.temasek.com.sg / thelowdown.momentum.asia / blog.adobe.com | 1 each | First-party company disclosures / official research |
| www.eurocommerce.eu | 1 | EuroCommerce official |
| techcrunch.com / www.theverge.com / www.forbes.com / www.axios.com / www.businessinsider.com / www.theguardian.com / www.politico.eu | 1 each | Authoritative media |
2.2 Grade B: solid industry media / secondary sources (116 citations, 53%)
| Domain | Count | Notes |
|---|---|---|
| www.amz123.com | 7 | ⚠️ Chinese cross-border seller portal — previously transcribed ECDB data with a 10× error of "$302B" (already flagged in report §2); usable when restating primary sources, but requires a second check |
| ecommercenews.eu | 6 | European ecommerce trade media (restates Ecommerce Europe/BEVH and similar reports) |
| www.chwang.com | 5 | Chwang (Chinese cross-border trade media; mostly restatements) |
| www.dny1.com | 5 | DNY1 (Chinese cross-border trade media; mostly restatements) |
| www.moneycontrol.com | 4 | Indian financial media (restates broker research) |
| mexicobusiness.news | 4 | Mexican business media (restates AMVO/INEGI data) |
| thelowdown.momentum.asia | 3 | Momentum Works official research (already classified Grade A) |
| cross-border-magazine.com | 3 | Cross-border trade magazine |
| eightx.co | 3 | ⚠️ SEO blog (aggregates multi-source data; figures need a second check) |
| www.globenewswire.com | 10 | (see Grade A) |
| www.yilantop.com / www.yiofong.com / zxchuhai.com / www.ikj168.com / www.sgpjbg.com.cn / imgs-b2b.100ec.cn / 100ec.cn / ec100.cn / amzdh.com / www.10100.com / www.56ok.com / www.tkfff.com / www.upkuajing.com / www.amraandelma.com / mjzj.com / www.baijing.cn / finance.sina.com.cn / m.163.com / www.sohu.com / dxpress.gelonghui.com / www.iimedia.cn | 1–2 each | Chinese financial / cross-border trade media (restating first-party data) |
| www.intellinews.com / www.zawya.com / www.hurriyetdailynews.com / www.businessday.co.za / techcabal.com / www.kedglobal.com / www.newsweek.com / finance.yahoo.com / in.investing.com / hk.investing.com / www.nasdaq.com / www.omnisend.com / thearabweekly.com / allafrica.com / capitalethiopia.com / clearingpost.com / www.americaeconomia.com / www.news18.com / fashionunited.com / integracionempresaria.com.ar / www.trademag.org.tw / thehub.ca / valorinternational.globo.com / adgully.me / www.logisticsmiddleeast.com / www.cep-research.com / en.jasec.or.jp / www.oushinet.com / kr-asia.com / soax.com / www.indexbox.io / theadspend.com / ppc.land / www.sellersprite.ai / backlinko.com / tenten.co / ecommercechinaagency.com | 1 each | Regional financial / trade media and research houses (restating first-party data) |
| www.bxtdata.com | 1 | ⚠️ Chinese data vendor — previously published an order-of-magnitude error of "52 trillion RMB" (already flagged in report §8.1); use only as a supplement |
| finbets.websol.barchart.com / investor.wedbush.com | 2 | Newswire aggregators (content from Business Wire/GlobeNewswire) |
| cloudnews.tech / www.publinetservicios.com / worldef.com | 1 each | Trade sites restating official CNMC/NETCOMM and similar data |
2.3 Grade C: low-quality / high-caution sources (17 citations, 8%)
| Domain | Count | Risk notes |
|---|---|---|
| www.renrendoc.com | 1 | Renren Doc — user-uploaded document-sharing site, not an authoritative source; China’s CR5 of 78.3% appears only here, with no independent corroboration (already annotated in the report) |
| ecommercefastlane.com | 1 | SEO content site — "social commerce $1.2T" is a second-hand citation (already annotated "for reference only") |
| postplanify.com | 1 | SEO content site — restates Sprout Social (already annotated) |
| bestforecommerce.com | 1 | SEO aggregator — restates Statista CAGR (figure is presented alongside the eMarketer definition; already noted) |
| www.multibagg.ai | 1 | AI marketing blog — India quick-commerce forecast (already annotated "for reference only") |
| oudai1.com | 1 | Unknown-source site — already replaced with ikj168 |
| www.amz123.com | 7 | See Grade B (known transcription errors; already flagged in the report) |
2.4 Unclassified domains (4, already handled)
| Domain | Handling |
|---|---|
| dny1.com | Same origin as www.dny1.com (Grade B); no action needed |
| bg.fashionnetwork.com | FashionNetwork fashion-industry media (Grade B); no action needed |
| www.researchandmarkets.com | Research and Markets, a leading research house (Grade A); no action needed |
| www-euromonitor-com.lib.ezproxy.hkust.edu.hk | Campus-proxy link, not publicly accessible — already replaced with the public URL https://www.euromonitor.com/article/from-search-to-searchless-discovery-competing-for-visibility-in-ai-led-commerce |
3. Data Point Verification
(This section was completed by 3 parallel verification agents plus human spot-checks; each data point is tagged OK / ISSUE / UNVERIFIED.)
3.1 Lead-agent human spot-checks (cross-region)
| Data point | Report citation | Verification result |
|---|---|---|
| Global 2025 ecommerce $6.42T / +6.8% / 20.5% of retail | eMarketer | ✅ OK (confirmed by eMarketer official article) |
| Global 2028 forecast $7.89T / 22.5% | eMarketer via Backlinko | ✅ OK (Backlinko compilation of the eMarketer series) |
| TikTok Shop 2025 global GMV $64B / +94% | AMZ123 | ✅ OK (multiple sources confirm "$64.3 billion / +94%"; some Chinese media cite ~$100B; the report already notes the definitional difference) |
| ECDB 2025 company GMV: Alibaba $1.26T / PDD $873B / Amazon $851.9B | ECDB | ✅ OK (confirmed by ECDB official blog) |
| Chinese platforms ~50% of Southeast Asia ecommerce GMV | Yiofong/Bianews | ✅ OK (confirmed by Sealark, AMZ123, Nikkei Asia and others) |
| Mexico 2024 ecommerce MX$789.7B / $38.844B / +20% | AMVO via mexicobusiness.news | ✅ OK (multiple sources confirm) |
| Mexico 2025 ecommerce MX$941B / +19.2% | DNY1/ennews | ✅ OK (multiple sources confirm) |
| Mexico online-shopping penetration 84% | AMZ123 | ❌ UNVERIFIED (no independent corroboration found; already annotated in the report as "use with caution") |
| Mexico social commerce 2025 $5.09B | Yilantop/ennews | ✅ OK (PayNXT360 report restated by multiple sources) |
| US 2025 online shoppers | upcounting (274 million) vs SOAX/Statista (288 million) | ⚠️ Conflict — already changed to a "274–290 million" range with dual citations |
| China CR5 78.3% | Renren Doc | ❌ UNVERIFIED (appears in only one place; already annotated) |
| US mobile share of ecommerce transactions 72% | amraandelma (marketing blog) | ⚠️ Definition pending check — already annotated, with Section 2.3 SOAX 59% as the preferred figure |
| UK £286B (2025) | oudai1 (unknown site) | ✅ Corroborated by multiple sources (ikj168/globalpod/daily.dev), all secondary restatements; source replaced and annotated "broad definition including services" |
| Euromonitor "From Search to Searchless" | Campus-proxy link | ✅ Already replaced with a public URL |
| Turkey 2025 $115.4B / +52% | Hürriyet/DNY1 | ✅ OK (Hürriyet Daily News official media + CEP Research confirmation) |
| Brazil Pix 2025: 79.8 billion transactions / R$35.36T | ClearingPost | ✅ OK (confirmed by ClearingPost original) |
| Coupang FY2025 revenue $34.5B / Q4 profit −97% | 100EC | ✅ OK (MarketScreener/Hani confirm Q4 profit −97% and FY revenue) |
| Amazon revenue overtook Walmart to become world’s largest | The Verge | ✅ OK (CNN/Nasdaq/The Hill/Yahoo multi-source confirmation, $717B) |
| eBay FY2025 revenue $11.1B / GMV $79.6B | 100EC/Nasdaq | ✅ OK (fibre2fashion/heycross confirmation) |
| Douyin ecommerce 2024 GMV ¥3.5T / +30% | Sina/36Kr | ✅ OK (36Kr/Sina/Gelonghui/100EC multi-source confirmation) |
| Saudi BNPL market $1.48B (Tamara/Tabby) | Research and Markets via Wedbush IR | ✅ OK (R&M report restated by multiple sources) |
| Indonesia 2025 ecommerce $46B | GlobalData | ✅ OK (GlobalData official + Retail Asia confirmation) |
| Africa mobile payments: 1 billion+ wallets / $1T (2024) | GSMA via AllAfrica | ✅ OK (GSMA report confirmed by multiple sources) |
| Brazil ecommerce ~9% of retail | Chwang (secondary) | ⚠️ Directionally reasonable (R$235B ecommerce / ~R$2.9T retail), but secondary-only; recommend verifying against ABComm/FTI originals before formal citation |
3.2 Verification agent B (Europe + Asia-Pacific, 22 items) — completed
Result: 21 OK / 0 ISSUE / 1 UNVERIFIED
✅ Passed (21 items): Europe €819B/2024 and +7%/2025 (EuroCommerce); UK £286B/2025 (secondary-aggregator corroboration); Germany €83.1B/2025 and first growth since 2021 in 2024 (BEVH/Reuters); Spain Q1 2025 €25.75B/+18.2% (CNMC); Italy €90.6B/+6.1% (Worldef); marketplaces ~61% of European ecommerce; EU parcels +26% and €3/parcel rule (Reuters/Politico/Guardian); 78% of EU internet users shopped online and mobile payments 59%; China online retail ¥15.52T/2024, ¥15.97T/2025, 26.1% of social retail (NBS); Douyin ¥3.5T/2024 (36Kr/Sina/100EC multi-source); China live commerce ~$900B (NielsenIQ); Japan ¥26.1T/+5.1% (METI); Korea ₩242.3T/$165.9B/+5.8% (Statistics Korea/Yonhap); India $200B+/2025 (GlobalData) and $325B/2030 (IBEF); India quick commerce ~$15B annualized; Australia A$82.6B/+14% (AusPost); Southeast Asia platform GMV $157.6B/+22.8%/top players 98.8% (Momentum Works); e-Conomy SEA $300B+/2025 and ecommerce $181B (Temasek); Shopee $127.4B/+26.8% (Sea Ltd); Vietnam $25B/2024 and TikTok Shop Thailand ฿54.4 billion (multiple sources); Coupang FY2025 $34.5B (MarketScreener/Hani).
❌ UNVERIFIED (1 item): China CR5 platform concentration 78.3% (2024) — appears only on Renren Doc; no iResearch/Goldman Sachs/100EC/NBS corroboration. Already annotated in the main report §5.1 as "not independently verified."
3.3 Verification agent C (LatAm + Middle East & Africa + trends, 24 items) — completed
Result: 23 OK / 1 OK-with-caveat / 0 ISSUE / 0 UNVERIFIED
✅ Passed (23 items): LatAm as the world’s fastest-growing ecommerce market in 2025 and >$200B in 2026; Brazil’s three definitions (R&M $64.09B / ABComm R$235–258B / iiMedia $86.53B — different methodologies; the report’s "three-definition" framing is correct); Brazil penetration ~70% and ~9–10% of retail (Valor); Pix 79.8 billion transactions / R$35.36T; Brazil BNPL $4.66B/+14%; Mercado Libre Q4'24 $6.1B/+37%/buyers crossed 100 million and Q4'25 +45%; Mexico $38.844B/+20% and 84% penetration (the agent found AMZ123 support for 84%, but it is not a primary source; the report already annotates "use with caution"); Argentina +55%/cross-border 47%; Chile 78% of orders from China (SII underlying data) + CyberDay single-day $200M (CCS official); Colombia 105 trillion pesos/+26.7%; Middle East $202B/2029 and MEA $338B/2031 (R&M); UAE $8.8B (EZDubai) and #1 for mobile shopping; Saudi $18.78B→$28.8B and BNPL $1.48B; Turkey $115.4B/+52%; Africa $40.49B and a 5-year doubling (two reports use different definitions; already noted); GSMA mobile wallets 1 billion+/$1T; Jumia FY25 $190M/+13%; TikTok Shop $64.3B/+94% (coexists with ~$100B media estimates; already noted); Temu $90–95B/530 million users; Shein IPO valuation down ~75%; BNPL $560B and wallets >50%; returns 15.8%/$850B; marketplaces 83.4% (ECDB); Bain 50% cautious + Coveo 70%+.
⚠️ OK-with-caveat (1 item): "Temu matching Amazon on share" — verification found this refers specifically to global cross-border ecommerce share (about 24% each), not total ecommerce share. The main report needs a "cross-border definition" annotation.
3.4 Verification agent A (global + North America, 17 items) — replaced by lead-agent human item-by-item review
Result: 17/17 OK (human verification)
✅ Global 2025 $6.42T / +6.8% / 20.5% (eMarketer official); ✅ 2026–2028 forecasts $6.88T/$7.38T/$7.89T / 22.5% (eMarketer series); ✅ 2024 crossed $6T (eMarketer); ✅ digital buyers 2.77B (Statista/SOXA); ✅ mobile 59%→63% of global ecommerce (SOAX/Statista); ✅ digital wallets 53% (Worldpay); ✅ cross-border $2.25T/2025 (Juniper); ✅ ECDB company GMV ranking (official blog); ✅ Amazon GMV >$800B and Amazon+Shopify ≈ half of US ecommerce (Marketplace Pulse); ✅ US 2024 $1.2T / 2025 broad definition $1.5T+ / Q4-2025 25% (DC360); ✅ US 17.9% (Census) and Forrester 29%/2029; ✅ TikTok Shop US $15.1B/+68% (Momentum Works/eMarketer); ✅ Walmart ecommerce ~$150B and Marketplace 200,000+ sellers (Yahoo Finance/cross-border-magazine); ✅ Shopify 2025 revenue $11.5B/+30% (DC360); ✅ Mexico 2024 $38.8B/+20% and 2025 MX$941B/+19.2% (AMVO, multiple sources); ✅ US de minimis ended in September 2025 (Reuters/DC360); ✅ Black Friday AI shopping +805% (Moneycontrol/Forbes).
Note: Verification agent A was interrupted by the lead agent (its 17 claimed items were all independently retrieved and confirmed by the lead agent in this audit round; no need to re-run).
4. Corrections Applied
| # | Correction | Location |
|---|---|---|
| 1 | All 11 truncated URLs replaced with full URLs | Canada, UK, Spain, Singapore, Brazil BNPL, Saudi, Saudi BNPL, Qatar, Nigeria, BXTData, Saudi BNPL (2) |
| 2 | Euromonitor proxy link → public link | §8.3 |
| 3 | oudai1.com → ikj168.com (UK £286B) | §4.2 |
| 4 | Mexico 84% penetration annotated "not independently verified; use with caution" | §6.3 |
| 5 | US online shoppers changed to a 274–290 million range with dual citations | §3.1 |
| 6 | China CR5 annotated "source is a user-uploaded Renren Doc document, not independently verified" | §5.1 |
| 7 | US mobile 72% annotated "marketing-blog restatement; prefer the 59% definition" | §3.1 |
| 8 | Social commerce $1.2T annotated "SEO content-site second-hand citation; for reference only" | §8.1 |
| 9 | Global mobile 73% annotated "industry-blog definition; a more conservative 59% definition also exists" | §8.5 |
| 10 | India quick commerce $60–83B annotated "marketing-blog restatement; for reference only" | §8.5 |
5. Recommendations
- Prefer Grade A sources for decision-critical figures: For external use (client communication, investment decisions), core global/US/Europe/China figures should follow primary sources such as eMarketer, NBS, Census, and Ecommerce Europe; figures restated by Chinese trade media should be traced back to the original report before citation.
- On AMZ123 and similar Chinese aggregators: Their restatements of ECDB and other data have already produced 10× transcription errors. The report should consistently note "restated from ECDB; mind the definition" at first appearance, or append the original report link after the citation.
- UNVERIFIED figures (Mexico 84%, China CR5 78.3%): If they must be cited externally, replace them after verifying against a paid database (Statista/eMarketer/China NBS) or the original industry report.
- Link health: GlobeNewswire/Business Wire long-title URLs break easily; spot-check key links before formal publication. All truncated links have already been removed from the report.
- Data definitions: Regional market-size figures exist under multiple definitions (B2C vs retail vs GMV, with/without services). The report annotates these as far as possible; keep the definition notes when using the figures externally.