Intake Triage: Cutting Response Time Without Adding Headcount

When a legal queue is slow, the instinctive diagnosis is capacity. In most departments we have looked at, the actual constraint is routing — work waits to be assigned far longer than it waits to be done.

Measure the wait, not the work

Break cycle time into three intervals: submission to acknowledgement, acknowledgement to assignment, assignment to completion. The third is the one everybody assumes dominates. Frequently the first two together account for most of the elapsed time, and neither requires additional lawyers to fix.

A structured front door

Free-text requests to a shared inbox guarantee a clarification round trip. A short intake form with conditional questions — matter type, counterparty, value, deadline, whether personal data or IP is involved — collects what is needed to route and prioritise on the first pass. Keep it under two minutes to complete; a form that takes ten minutes will be bypassed.

Route by rule, escalate by exception

Most requests fall into a handful of types with an obvious owner. Encode that as routing rules and reserve human triage for what does not match. Departments that do this typically cut assignment time from days to minutes for 70–80% of intake.

Publish the service levels

Different response times for different urgency levels, published and visible. This is not primarily a promise to the business — it is a mechanism for making the business declare urgency honestly, which is what allows prioritisation to work.

Deflect what does not need a lawyer

Track the questions that arrive repeatedly. A guidance page, a template with instructions, or a self-service NDA flow with automatic escalation on non-standard answers removes whole categories from the queue. In our experience, 20–30% of intake in a typical commercial legal team is answerable this way.

What this achieves

Faster response, better prioritisation and a clean dataset about what the business actually asks for — which is the evidence needed if headcount really is the answer.