JSON data quality guide

How to Inspect JSON Data Quality Before Import or Integration

Valid JSON can still contain missing properties, changing field types, inconsistent object shapes, duplicate paths, and identifiers that are not unique. A repeatable inspection sequence exposes those risks before the document reaches an API, database, spreadsheet, or analytics pipeline.

Open JSON Inspection Summary Report

Start with syntax, then inspect the document as data

Syntax validation answers whether a document can be parsed. Data-quality inspection asks different questions: whether repeated records share fields and types, how complete those fields are, which values repeat, and how much the payload costs to move.

Run the JSON Inspection Summary Report first when you need a broad baseline. It combines the established type, completeness, frequency, structure, size, duplicate-path, structural-consistency, and cardinality reports without changing the source.

  1. Paste or upload the original JSON document.
  2. Confirm that syntax validation succeeds.
  3. Generate the consolidated inspection report.
  4. Save the JSON report with the source version or QA record.

Measure completeness and type stability

A field can be present, explicitly null, an empty string, or missing entirely. Those states often have different downstream meanings, so the completeness report measures them separately.

Type profiling shows the value types observed at normalized paths. Review paths that mix strings, numbers, booleans, objects, arrays, or null before an importer infers a column type.

Before

[{"id":101,"active":true},{"id":"102","active":null},{"id":103}]

After

Review: id changes from number to string; active is null once and missing once.

Compare object shapes and field cardinality

Structural consistency compares records with the dominant observed field set and types. It reports missing fields, extra fields, and type mismatches by object, making schema drift visible without generating or enforcing a schema.

Cardinality measures distinct values as a percentage of present values. Low-cardinality fields often behave like categories, while complete string or number fields with one value per record are reported as candidate identifiers. Confirm business meaning before treating any candidate as a key.

Review paths, size, and the handoff boundary

Duplicate normalized paths explain how often locations recur through arrays, while structure and size reports show nesting, container counts, payload bytes, and estimated minification savings.

Use the focused analyzers when a summary finding needs detail. Resolve or document important variations, then validate the final document against the receiving system's contract before import or integration.

  1. Investigate missing, extra, or mixed-type fields.
  2. Confirm candidate identifiers against domain rules.
  3. Review payload size and deeply nested structures.
  4. Re-run the report after approved source corrections.
  5. Validate against the destination schema or import requirements.

Common questions

FAQ

Is valid JSON automatically high-quality data?

No. Syntax validity does not guarantee complete fields, stable types, consistent record shapes, unique identifiers, or appropriate values.

Does UtilityDock infer or enforce a schema?

The Inspection tools report observed measurements. Use JSON Schema Validator separately when you have a formal schema to enforce.

What does 100% field cardinality mean?

Every present value is distinct. If the field is also complete and contains only strings or numbers, UtilityDock reports it as an identifier candidate.

Does the inspection report upload my data?

No. Your files never leave your computer. All inspection and report generation happens locally inside your browser.