The word “native review” gets sold as a quality tick-box. Here is what a good native reviewer is really looking for, and why it is the step that catches the errors that matter.
- Grammar is the part a checker already caught
- Register: correct words, wrong room
- The things that are “right” but land wrong
- How to tell a real native review from a rubber stamp
Grammar is the part a checker already caught
"Native review" is on every vendor’s quote, which has quietly turned it into a box people tick without asking what it means. If all it delivers is a fluent speaker fixing grammar, you are paying a premium for something automated tools already flag. Grammar and spelling are the floor. A native review that stops there is not worth its line item.
What a good native reviewer is actually for is the layer beneath grammar — the things that are technically correct and still wrong for the market. That is where the value is, and it is invisible on a quote.
Register: correct words, wrong room
The most common real defect a native reviewer catches is register — the level of formality and tone. A translation can be flawlessly accurate and pitched at completely the wrong level for its reader: a consumer safety warning written in stiff legal register that nobody reads to the end, a marketing line rendered so literally it sounds like a manual, a formal filing that drifts casual in one paragraph.
A non-native speaker, however fluent, tends not to feel this. They can confirm the sentence is correct. They cannot always feel that it sounds like a lawyer wrote it when it should sound like a brand, because that judgement comes from living inside the language, not studying it.
The things that are “right” but land wrong
A native reviewer is also the last defence against the renderings that are defensible in a dictionary and wrong in the market: a phrase that is technically fine but carries an unintended connotation locally, a number or date format that is correct-but-foreign, an idiom translated so faithfully it becomes nonsense, a term that is standard in one region and unknown in the target one.
None of these are errors a QA tool flags, because none of them are errors in the narrow sense. They are the difference between text that was translated and text that belongs. That difference is the entire reason native review exists.
How to tell a real native review from a rubber stamp
Ask what the reviewer changed and why. A genuine native review produces edits with reasons — "shifted register, the source is a safety notice not a contract"; "localised the example, the original reference means nothing in this market". A rubber stamp produces a clean file and a green tick with nothing to show.
And ask where the reviewer is. Native review by someone who left the target market fifteen years ago is weaker than it sounds — language and usage move, and for anything consumer-facing you want a reviewer who is still inside the culture, not merely fluent in its language. On regulated or technical content, add the second requirement: native and domain-qualified, because feeling the register does not help if you cannot also judge the term.