Short answer
Law firms should separate fast intake from legal judgement. Capture the enquiry quickly, protect confidentiality, run the right checks, route the matter and keep follow-up visible.
What the calculator is showing
According to the 2025 Professional Services Client Journey Report by insight6, law firms scoring below 80% on enquiry handling are materially less likely to convert opportunities. For a firm handling 100 enquiries a month, the model below uses a 93% at-risk enquiry rate, a 30% conversion drop and a £4,000 average client value. That is how the annual leakage figure reaches about £1.34m.
Speed matters, but not at any cost
A conveyancing, private client, employment or commercial enquiry can go cold quickly. The first response should make the client feel heard and tell them what happens next. It should not create legal advice, accept the matter or promise outcomes before the right checks are complete.
Protect the intake boundary
The intake workflow should collect enough information to route the enquiry, but not invite sensitive detail before the firm is ready to handle it. Keep conflict checks, confidentiality warnings and matter acceptance rules clear. If automation is used, it needs strict limits on what it asks and stores.
Make ownership visible
A new enquiry needs an owner, even if the fee earner is in court, in meetings or buried in billable work. The workflow should show who received it, who is responsible for the next action, and when the prospect should hear back.
Follow-up should not rely on memory
Many law firms lose prospects after the first acknowledgement. Build follow-up around clear statuses: new, awaiting conflict check, awaiting fee earner review, call booked, proposal sent, closed. Each status should trigger the next sensible action.
Use AI carefully
AI can summarise intake notes, draft safe acknowledgements, route matters by practice area and flag missing information. It should not provide legal advice, assess prospects as clients or handle confidential documents without an approved control model.