diff --git a/services/search/analytics.py b/services/search/analytics.py
index 39b00dd..64e61e9 100644
--- a/services/search/analytics.py
+++ b/services/search/analytics.py
@@ -45,32 +45,36 @@ class RateLimitError(SearchEngineError):
 # ----------------------------------------------------------------------
 # Analytics helpers
 # ----------------------------------------------------------------------
+def _default_analytics() -> Dict[str, Any]:
+    return {
+        "total_queries": 0,
+        "successful_queries": 0,
+        "failed_queries": 0,
+        "cache_hits": 0,
+        "cache_misses": 0,
+        "query_patterns": {},
+    }
+
+
 def _load_analytics() -> Dict[str, Any]:
     """Load analytics data from the JSON file, creating defaults if missing."""
     if not ANALYTICS_FILE.exists():
-        default = {
-            "total_queries": 0,
-            "successful_queries": 0,
-            "failed_queries": 0,
-            "cache_hits": 0,
-            "cache_misses": 0,
-            "query_patterns": {},
-        }
+        default = _default_analytics()
         _save_analytics(default)
         return default
     try:
         with open(ANALYTICS_FILE, "r", encoding="utf-8") as f:
-            return json.load(f)
+            data = json.load(f)
+        # Merge over defaults so a file written by an older schema (or a
+        # partial write) still has every counter — _record_query indexes
+        # these keys directly and would otherwise raise KeyError.
+        merged = _default_analytics()
+        if isinstance(data, dict):
+            merged.update(data)
+        return merged
     except Exception as e:
         logger.warning(f"Failed to load analytics file: {e}")
-        return {
-            "total_queries": 0,
-            "successful_queries": 0,
-            "failed_queries": 0,
-            "cache_hits": 0,
-            "cache_misses": 0,
-            "query_patterns": {},
-        }
+        return _default_analytics()
 
 
 def _save_analytics(data: Dict[str, Any]) -> None:
diff --git a/tests/test_services_search_analytics_defaults.py b/tests/test_services_search_analytics_defaults.py
new file mode 100644
index 0000000..a0a67c2
--- /dev/null
+++ b/tests/test_services_search_analytics_defaults.py
@@ -0,0 +1,41 @@
+"""Default-merge on load for services/search/analytics.py.
+
+src/search/analytics.py was fixed to merge a loaded analytics file over
+defaults so _record_query never hits a missing counter, but the services
+copy diverged and still returns json.load(f) verbatim. The services copy
+is the live one: services/search/core.py calls _record_query on every
+search, so an analytics file missing a key (older schema or partial
+write) raises KeyError and breaks comprehensive_web_search.
+
+Mirrors tests/test_search_analytics_defaults.py which covers the src copy.
+"""
+import json
+
+import services.search.analytics as analytics
+
+
+def test_load_merges_defaults_for_partial_file(tmp_path, monkeypatch):
+    f = tmp_path / "search_analytics.json"
+    f.write_text(json.dumps({"total_queries": 5}), encoding="utf-8")
+    monkeypatch.setattr(analytics, "ANALYTICS_FILE", f)
+
+    data = analytics._load_analytics()
+
+    assert data["total_queries"] == 5
+    assert data["query_patterns"] == {}
+    for key in ("successful_queries", "failed_queries", "cache_hits", "cache_misses"):
+        assert data[key] == 0
+
+
+def test_record_query_survives_partial_file(tmp_path, monkeypatch):
+    f = tmp_path / "search_analytics.json"
+    f.write_text(json.dumps({"total_queries": 1}), encoding="utf-8")
+    monkeypatch.setattr(analytics, "ANALYTICS_FILE", f)
+
+    # Before the fix this raised KeyError on the missing counters.
+    analytics._record_query("hello world", success=True, cache_hit=False)
+
+    data = analytics._load_analytics()
+    assert data["total_queries"] == 2
+    assert data["successful_queries"] == 1
+    assert data["query_patterns"]["hello world"]["count"] == 1