Company Registry CSV Import:
Preserve Leading Zeros & UTF-8.
.csv directly in Microsoft Excel—automatic numeric casting strips leading zeros from 8-to-14-digit identifiers (CRN, KvK, KRS, SIREN, CNPJ) and turns long IDs into irreversible E+12 floats. Either open our pre-typed .xlsx workbooks directly or import CSVs via Excel Power Query (65001: UTF-8, Text columns), dtype=str in Python, or VARCHAR in SQL.
High-Risk National Registry Identifiers Vulnerable to Excel Auto-Casting
| Jurisdiction | Identifier / Field | Raw UTF-8 CSV Value | Broken Excel Double-Click | Required Schema Type |
|---|---|---|---|---|
| United Kingdom → | Companies House CRN (8 chars) | "01234567" |
1234567 (7 digits — broken join) |
Text / VARCHAR(16) · Pad 8 chars |
| France → | INSEE SIREN (9d) / SIRET (14d) | "05480012300018" |
5.48001E+12 (scientific notation) |
Text / VARCHAR(18) · Never float |
| Netherlands → | KvK-nummer (8 digits) | "08094512" |
8094512 (leading 0 stripped) |
Text / VARCHAR(12) · LPAD 8 |
| Brazil → | Receita Federal CNPJ (14 digits) | "00394460005887" |
3.9446E+11 (checksum lost) |
Text / VARCHAR(18) · Keep 14 digits |
| Poland → | KRS (10 digits) / REGON (9/14d) | "0000145732" |
145732 (4 leading zeros lost) |
Text / VARCHAR(14) · LPAD 10 |
| Norway → | Organisasjonsnummer + Postal (4d) | Postal "0150" (Oslo) |
150 (invalid Norwegian postcode) |
Text / VARCHAR(10) on postal_code |
| Indonesia & Malaysia → | NIB (13 digits) / SSM (12 digits) | "0220104567891" |
2.20105E+12 (truncated precision) |
Text / VARCHAR(20) · UTF-8 |
How Do You Import Corporate CSV Files in Excel Power Query Without Losing Zeros?
Importing corporate CSV files in Excel Power Query preserves both UTF-8 diacritics (German umlauts, Polish/Lithuanian characters, Nordic letters, French/Spanish accents) and leading zeros in five steps: open a blank workbook, click Data → From Text/CSV, select 65001: Unicode (UTF-8) as File Origin, click Transform Data, highlight entity_id, registration_number, postal_code, and phone to set Data Type to Text, and click Close & Load. For files exceeding Excel's 1,048,576-row ceiling (such as United States, Germany, or United Kingdom), choose Close & Load To → Only Create Connection (Add to Data Model / Power Pivot) or use Python/DuckDB.
Reference Ingestion Snippets (Python Pandas & DuckDB SQL)
# Python (pandas) — Zero-loss string ingestion + SimHash / domain deduplication
import pandas as pd
df = pd.read_csv(
"Netherlands-Companies-Database-OpenData.csv",
dtype=str, # Preserves 08094512 KvK, postal codes & phone prefixes
encoding="utf-8",
keep_default_na=False
)
# DuckDB — Instant query over 1M+ rows without loading into RAM
# SELECT * FROM read_csv('Germany-Companies-Database-OpenData.csv', all_varchar=true) LIMIT 50;
How Should HubSpot and Salesforce Map Open Company Data Standard v1.1 Columns?
When syncing Open Company Data Standard v1.1 exports into HubSpot or Salesforce, map website_domain as the primary Account/Company deduplication key alongside a custom External ID property storing iso_country + ':' + registration_number (documented in our Registry Identifiers Guide). Test your CRM mapping first with any free 50-row CSV sample from Switzerland, France, Brazil, or Indonesia.
Frequently Asked Questions
Why does Microsoft Excel strip leading zeros from SIREN, KvK, CRN, and CNPJ codes when opening a CSV?
Double-clicking a plain-text CSV triggers Excel automatic type inference, which casts fixed-width numeric strings (like UK Companies House 01234567, French SIREN 054321987, Dutch KvK 08123456, or Brazilian CNPJ 00123456000199) into 64-bit integers/floats and drops leading zeros or converts 14-digit codes to scientific notation.
What is the fastest zero-loss way to open Central.Enterprises files in Excel?
Either download our native .xlsx Excel edition directly (where identifier and postal_code columns are pre-typed as Text), or import the UTF-8 .csv via Excel Data -> From Text/CSV (Power Query), setting File Origin to 65001: Unicode (UTF-8) and typing entity_id, registration_number, and postal_code as Text.
How should Python pandas, DuckDB, or PostgreSQL read corporate registry CSV files?
Always force string/VARCHAR types on identifier, phone, and postal_code columns at read time—for example pd.read_csv(path, dtype=str, encoding="utf-8") in Python or all_varchar=true in DuckDB—before casting numeric metrics.