Null density is the percentage of fields that have a null value. High null density (20%+) usually means the API is returning fields that don’t apply to this resource — for example, returning optional metadata fields as null for every record instead of omitting them. Fields with null values still consume bandwidth and parse time.

JSON is highly repetitive — field names appear once per object in arrays, and common values repeat. GZIP compression typically reduces JSON size by 70–80% in transit. Enable it with the Content-Encoding: gzip header. Most web servers and CDNs support it automatically when configured.

Over-fetching means the API sends more data than the client needs. A common example: a mobile app showing a user’s name and avatar receives a response with 30+ fields including full address history and settings. GraphQL was invented specifically to solve this — clients request exactly the fields they need.

Beyond 3–4 levels of nesting, JSON becomes hard to read, hard to document, and slow to traverse in code. Deep nesting often signals that the API should be broken into separate endpoints or that the data model needs rethinking. This tool flags responses with nesting beyond 4 levels.