The firm that stops losing clients at 8pm.
Clio Legal Trends Report 2024 — a third-party secret shopper contacted 500 US law firms with the questions a prospective client would ask. US data; no equivalent Canadian study has been published, and we have not yet run one.
The situation this is built for
A small family law practice with strong referrals and good reviews, and an intake process that only works during business hours. People dealing with separation and custody research at night — and the research above suggests roughly half of firms are reachable by neither phone nor email when they do. Many firms have also bought a website chat widget that collects names and goes nowhere, which is a different problem wearing the same clothes.
The scope — one workflow, one number
The number to move: booked consultations per month. The workflow: first contact to booked consult. Nothing else is touched in phase one — that restraint is the reason it takes two weeks instead of two months.
- 01
24/7 intake agent on the existing number and website — answers instantly, explains the consultation process, and books directly into each lawyer's calendar. It gathers the practical facts (matter type, urgency, opposing party name for conflict screening) and never gives legal advice. That boundary is configured and reviewed by you before launch, not left to the model's judgment.
- 02
Conflict-aware handoff — every intake lands in the CRM with the conflict-screening fields filled, flagged for a human check before the consult is confirmed. Anything sensitive routes straight to staff with the full conversation attached.
- 03
Phase two, once phase one has earned it: document chasing — retainers, financial disclosure checklists and ID requests following up on their own schedule rather than a clerk's. We do not build this in phase one, because if the consult number does not move there is no point automating what happens after it.
Privacy, by design
Family law enquiries are sensitive. The agent collects the minimum needed to book and screen, data stays in the firm's own accounts with Canadian residency where the underlying tool offers it, nothing trains any model, and the firm holds a written data map at handover — the documentation an insurer tends to ask for and rarely receives.