This article documents how to extract country-level, industry-level value-added data from UNIDO’s industrial statistics database “INDSTAT.”

No official API documentation is published, but data can be retrieved using the browser’s internal API.

Background

When researching overseas food market sizes or manufacturing sector scales, UNIDO INDSTAT is one of the most reliable data sources available.

However, the official site (stat.unido.org) is protected by Cloudflare security, making standard scraping impossible.

This article records the method that actually worked.

What Is UNIDO

UNIDO (United Nations Industrial Development Organization) is a UN agency that collects and publishes manufacturing statistics from countries worldwide. INDSTAT is a database organizing manufacturing data from 170+ countries by industrial classification (ISIC).

Item Details
Official Name UNIDO INDSTAT Rev.4
URL https://stat.unido.org
Classification ISIC Rev.4 (2-digit codes: Food=10, Chemicals=20, Motor Vehicles=29, etc.)
Available Indicators Value added, output, employment, number of establishments, wages, capital investment
Country Coverage 170+ countries
Cost Free (no registration required)
Update Frequency 1–2 times per year

Three Barriers to Data Retrieval

Retrieving UNIDO data involves three hurdles: Cloudflare bot protection, undocumented API specifications, and missing China data.

Each is explained below.

Barrier 1: Cloudflare Bot Protection

stat.unido.org is protected by Cloudflare Turnstile.

Standard HTTP requests (requests, curl, etc.) cannot access it.

Solution: Use a browser automation tool (Playwright) with stealth mode (playwright-stealth) to pass Cloudflare, then call the API from within the browser.

Barrier 2: Undocumented API

UNIDO does not publish official API documentation.

The endpoint and parameters must be identified by analyzing requests in the browser’s Developer Tools (Network tab).

Below are the API specifications discovered through analysis.

Endpoint Information:

Item Details
Endpoint https://stat.unido.org/portal/dataset/getData
Method POST
Content-Type application/json

Request Parameters:

Parameter Example Value Description
datasetId 125 INDSTAT Rev.4 ID (Rev.3 is 110)
datIns “25 Sep 2025” Dataset last update date
countryCode “764” ISO 3166-1 numeric (3-digit). Thailand=764
periods [] Empty array retrieves all periods
variableCode “20” Indicator to retrieve. 20=Value Added

variableCode List:

Code Indicator
01 Number of establishments
04 Number of employees
05 Wages and salaries
14 Output
20 Value added
21 Gross fixed capital formation
31 Female employees

Response Fields:

Field Description
p Year (e.g., “2021”)
a ISIC code (“10”=Food manufacturing, “C”=Total manufacturing)
v Value in local currency
u Value in US dollars

Barrier 3: Missing China Data

China does not submit ISIC-classified data to UNIDO.

This is because China uses its own industrial classification (GB/T).

To obtain China’s food manufacturing data, manual retrieval from the National Bureau of Statistics (data.stats.gov.cn) with USD conversion is required.

Key Country Codes

Country codes use ISO 3166-1 numeric (3-digit numbers). Note: these are NOT the 2-letter ISO codes (JP, TH, etc.).

Code Country Code Country Code Country
392 Japan 764 Thailand 840 United States
156 China (no data) 704 Vietnam 276 Germany
410 South Korea 360 Indonesia 826 United Kingdom
356 India 608 Philippines 250 France
076 Brazil 458 Malaysia 484 Mexico
158 Taiwan 702 Singapore 643 Russia

ISIC Rev.4 Major Industry Codes (2-digit)

Food is ISIC 10. Chemicals is 20. Motor vehicles is 29. Using the wrong industry code returns completely different data.

Code Industry
10 Food products
11 Beverages
13 Textiles
17 Paper and paper products
20 Chemicals
21 Pharmaceuticals
22 Rubber and plastics
24 Basic metals
26 Computer, electronic and optical products
27 Electrical equipment
29 Motor vehicles

Retrieval Procedure

The only required libraries are playwright and playwright-stealth.

Steps:

  • Access stat.unido.org with Playwright and pass Cloudflare (wait approximately 8 seconds)

  • Call the API via in-browser JavaScript execution (page.evaluate)

  • Extract data from the JSON response where ISIC code = “10” (food) for the latest year

  • Loop through multiple countries

Key Points:

  • Execute fetch via page.evaluate inside the browser. Use the browser’s session cookies, not Python’s requests library

  • Once past Cloudflare, the API can be called repeatedly within that browser session

  • There is no bulk-download API. Requests must be made one country at a time

Summary

Item Details
Data Reliability High (collected by a UN agency from national governments)
Retrieval Difficulty Moderately high (requires Cloudflare bypass and API reverse-engineering)
Limitations No data for China, Cambodia, or Myanmar
Alternative Manual search on the website (easier for small volumes)
Update Check New data is available when the datIns parameter date changes

Sources

UNIDO Database:

Industry Classification:

China Data Supplementation:

Technical (Cloudflare Bypass / Browser Automation):

Related International Statistical Databases:

About the Researcher

Takashi Kinoshita — CEO, Taitonmai Co., Ltd.

  • M.A., a national university graduate school

  • 8 years in procurement at SHARP Corporation

  • Including 2 years stationed at SHARP’s Thailand factory as Procurement Manager, leading local staff

  • Sole Japanese manager on-site; conducted procurement operations in English and Thai

  • After independence, completed corporate research covering 80+ countries and 10,000+ companies

Research Date: February 11, 2026 (data retrieval) / February 14, 2026 (article creation)

For overseas company list research and market sizing → https://taitonmai.co.jp