From a3fa6383ed91b0cb71a4cef61a8940af76e6111d Mon Sep 17 00:00:00 2001
From: bk86a <41694587+bk86a@users.noreply.github.com>
Date: Wed, 25 Feb 2026 15:42:26 +0100
Subject: [PATCH] docs: add 99.3% success rate from 134M postal code bulk test

---
 README.md | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

diff --git a/README.md b/README.md
index 2d651af..38dba50 100644
--- a/README.md
+++ b/README.md
@@ -19,7 +19,9 @@ North Macedonia (MK), Serbia (RS), Türkiye (TR)
 
 ## Testing
 
-The service has been tested against **134 million real-world postal codes** from 34 countries, sourced from 8 publicly available European datasets: GeoNames, GLEIF, TED, SIRENE, OffeneRegister, FTS, OpenAddresses, and Erasmus+ ECHE. All are open data published under permissive licenses (CC BY 4.0, CC0, or Licence Ouverte v2.0).
+The service has been tested against **134 million real-world postal codes** from 34 countries, sourced from 8 publicly available European datasets (GeoNames, OpenAddresses, GLEIF, SIRENE, TED, OffeneRegister, FTS, and Erasmus+ ECHE). All are open data published under permissive licenses (CC BY 4.0, CC0, or Licence Ouverte v2.0).
+
+After deduplication, **970,083 unique postal codes** were tested with an overall success rate of **99.3%**. The remaining failures are predominantly data quality issues in the source datasets (placeholders, cross-country codes, legacy formats) rather than gaps in NUTS coverage.
 
 ## Quick start