What LQA Is, and What It Costs You to Skip It

What LQA actually is

LQA means playing the game in the target language, on a real build, with a native speaker at the controls. The text has already been translated and edited by this point. What LQA checks is what happened to that text once it went into the game.

That distinction is the whole point. Most localization bugs are not translation errors. The words can be perfect in the spreadsheet and still be wrong on screen: clipped by a button, rendered as empty boxes, attached to the wrong speaker, or grammatically correct for a man when the player is playing a woman.

Why it is worth paying for separately

A translator working from a spreadsheet is working blind. They see a string, a character limit if they are lucky, and a note if the developer wrote one. They do not see the button the string sits on, the character who says it, or the three other strings it appears next to.

Everything that goes wrong in that gap is invisible until someone runs the game. Nobody on the development side can catch it either, because reading Korean menus is not part of a producer’s job. LQA is the only stage where a person who can read the language and a build that actually runs are in the same room.

What LQA catches

Roughly in order of how often we file them:

What happens when you skip it

The bugs do not disappear. The finder changes. Instead of a tester filing a report with a screenshot and a string ID, you get a player posting the screenshot with a caption, in a language your community manager does not read.

Three costs follow, and only the first one is obvious.

1. The fix gets more expensiveBefore release, a truncated button is a line in a report. After release, it is a patch, a build, a submission, and a changelog.
2. The reviews are permanentStore pages keep localization complaints at the top for years, and they are the first thing the next player in that market reads.
3. Certification can bounce the buildFirst-party naming rules are not advisory. A build can fail submission over terminology that a tester would have flagged in an hour.

Why LQA is a good deal

LQA is billed by the hour, not by the word. That makes it one of the smallest lines in a localization budget and the one with the widest reach: a tester covers every string that renders, including the ones nobody remembered to send for translation.

It is also the only stage that tests the pipeline itself. A single LQA pass tells you whether your export was complete, whether your fonts cover the language, whether your UI survives a longer word. Those answers keep paying off on every update you ship afterwards.

How we run LQA

Our testers are native speakers who play games in their own language for fun, which is the only reliable way to know how the language actually sounds in a game rather than in a document.

We never put a translator on their own text. Somebody who wrote a line reads back what they meant by it, not what is actually on the screen. A tester who meets the build cold is in exactly the position your player will be in, and that is the position worth testing.

What does carry over is the project knowledge. The tester gets the glossary, the style guide and the terminology decisions already made on the project, so they inherit the context without inheriting the blind spots.

You get a structured report: what broke, where, in which language, with a screenshot and a string ID. We fix the text-side issues ourselves. Anything that needs the engine goes back to your team with steps to reproduce it, so nobody spends an afternoon hunting for a bug that takes four button presses to reach.

Our native speakers play-test the game before other native speakers do. The difference between ours and the others is simple: ours send you comprehensive bug reports, rather than go online to write potentially harmful reviews.