Clearing the batch teaches a firm nothing about judgment

The pilot passes and the real question is still open

A passed review pilot answers one question: can this team apply a manual at volume. The decision sitting behind it is a different one entirely. Whether offshore staff can hold a seat with a decision in it. Intake, case management, drafting, client updates, calendar responsibility. None of that is graded on agreement with a second pass. Review results do not transfer to that decision, and they were never measuring the same thing. So the pilot passes and the hire nobody has made yet is still unmade. The batch cleared, which is production and not evidence.

Then the work vanishes. A matter closes and the review closes with it. No standard operating procedure survives it. No repository grows. Nobody offshore understands the firm better in month six than in month one, because nothing was written down that outlives the project. The next matter starts from zero, with the same vendor and the same manual. And a team with no next matter has nothing to stay for. The firm bought throughput on one matter and called it a staffing test. Months of production, and nothing on the bench a second matter could use.

A pile that moved is not a bench.

How much does offshore document review actually save?

First-pass coding gets cheaper offshore. The attorney hour sitting behind it does not move. Privilege calls, close questions and the final read stay with a licensed attorney, priced as they were before the vendor arrived. The ABA Model Rules set supervision obligations for non-lawyer assistants. Delegating the first pass does not lift them. Where the assistant sits changes nothing about who answers for the output. So the per-document quote is describing half the job.

Which is why the vendor's rate card is the wrong document to argue over. Outsourcing changes the cost of producing work. It does not change the cost of reviewing it. Review is the input most firms are already short of. The per-document rate is the only thing on the record, and it is not the number that decides. Count the weekly hours the named reviewer can give the matter first. Price the second pass, or the project has not been priced.

What to send offshore first instead of the review backlog

A better first job is one where being wrong is visible and recoverable, not one where being wrong is impossible. Three tests separate the two, and I would run them before anyone is interviewed.

• Does the work still exist after this matter closes? If not, the pilot buys a result and no capability.
• Is there a written standard for a correct output, or does it live in one partner's head? Criteria nobody wrote down cannot be handed to anybody.
• Will the work generate escalations? A first job that produces no questions has tested nothing worth knowing.

Intake summaries, medical record chronologies, demand package assembly, calendar and deadline tracking. Each one recurs weekly. Each has a correct shape somebody can write down, and each produces a stream of judgment calls a supervisor can read. Run the trial against a matter the firm already closed, so it grades the provider instead of grading the firm's own uncertainty. Then read what comes back with the attention a filing gets. Review never gave anybody a reason to read it that closely.