A Professional Services Firm: Reducing Manual Intake Work with an AI Assistant

Challenge & Objectives

The firm’s partners and senior associates were personally screening every inbound contact-form submission to determine whether it matched a practice area they handled, whether it was a fit, and who internally should take it — work that had nothing to do with billable client service but consumed real hours every week. The objective was to remove that manual screening burden without losing good-fit inquiries in an automated system.

Research

A review of several months of inbound inquiries found that a large share were either outside the firm’s practice areas entirely, or were a good fit but got delayed because the right team member wasn’t looped in until someone manually forwarded the message — sometimes days later. The firm had no structured intake questions; the contact form was a single open text box.

Strategy

The strategy replaced the single open text box with a short structured intake flow, backed by an AI assistant that asks a small number of clarifying questions based on the visitor’s initial description, classifies the inquiry by practice area and urgency, and routes it directly to the right team member with a concise summary already prepared. The mechanism: most of the “screening” work partners were doing by hand was really just gathering the same handful of facts (practice area, situation summary, timeline) every time — work well suited to being asked once, consistently, by an assistant, rather than repeatedly by a senior person.

AI & Web

The assistant was built into the firm’s existing “Contact Us” and practice-area pages as a conversational intake flow rather than a separate tool visitors had to seek out. Anything genuinely ambiguous, sensitive or outside a clear practice-area match is flagged for direct partner review rather than the assistant attempting to force a classification.

Automation

Completed intakes sync directly into the firm’s CRM with the structured summary attached, and the relevant team member is notified immediately rather than the inquiry sitting in a shared inbox — closing the multi-day gap that had previously delayed even strong-fit inquiries.

Implementation

We ran the assistant alongside the old contact form for several weeks, comparing intake quality and completion rates before fully replacing the form, and refined the clarifying questions based on where visitors most often needed extra prompting to describe their situation clearly.

These are cited, published industry benchmarks used as context for the methodology — not a verified, client-approved outcome for this anonymized case study.

• Manual intake screening is frequently repetitive fact-gathering, not judgment work — a good candidate for automation, with a clear human-review fallback.
• A structured intake flow surfaces better information than a single open text box, for the visitor and the firm alike.
• Routing speed matters as much as routing accuracy — a good-fit inquiry that sits unrouted for days can still be lost.