Definition

The Review Gap

The Review Gap is the unaddressed half of accountancy automation: a decade of investment went into preparation (bookkeeping, bank feeds, coding, draft accounts) while review, the conversion of prepared work into signed work, was left as one senior human reading by hand.

Why it matters

The gap is where the trouble in a practice actually lives. Lock-up lives there, because files queue for the one person who can sign them. Write-offs live there, because problems found late are problems absorbed rather than billed. Quality variation lives there, because a tired reviewer on a Friday is a different instrument from a fresh one on a Tuesday. The partner bottleneck is not a diary problem; it is the Review Gap wearing a diary as a disguise.

How it closes

Not by trusting a model to sign, and not by adding another checklist to the same overloaded human. It closes by moving the labour of review to the machine and the judgement to the person: every material figure carried up the ladder from CLAIMED to PROVEN automatically, so the human hours land on the findings that need a decision. That is the Trace-to-Source Standard, and it is why evidence-led AI accountancy treats review, not preparation, as the point of the technology.

Common questions

Is the Review Gap an AI problem?

No, it predates the current models by years. AI simply makes it urgent: production is now cheap enough that unreviewed output can be generated faster than any senior can read it.

Does closing it mean fewer reviewers?

It means better ones. The hours move from hunting for evidence to weighing it, and trainees learn judgement earlier because the grind is gone. We did not lose anyone to AI. We moved them up to it.

What should a firm measure?

Lock-up days, write-offs on review-stage problems, and the spread between the best and worst reviewed file in a month. All three narrow as the gap closes.

The mechanics are in the Trace-to-Source Standard and the full operating model is free in The Signed Review.