How a recruitment agency qualified 3x more leads with an AI receptionist

The problem

The agency runs eight specialist desks across tech, finance and life sciences. Inbound enquiries arrived from job boards, LinkedIn campaigns, the website chat widget and the main phone line, and every consultant was expected to triage their own pipeline.

The triage burden had become unmanageable. Each consultant was fielding 60 to 90 first-touch enquiries a week, of which only one in five was a genuinely qualified candidate or a live client brief. The other 80% were CV blasts, recruiters fishing, or vendors selling.

Speed-to-lead had collapsed. The agency own data showed an average first response time of 4.2 hours during business hours and effectively never out of hours, despite their largest competitors responding within minutes on LinkedIn.

Two attempts at outsourced first-touch had failed: agents did not understand industry vocabulary, could not pre-qualify against the client brief, and routinely escalated time-wasters as if they were real opportunities. The managing director needed first-touch that was fast, fluent and surgical about who it escalated.

The solution

Boafo Agent was deployed on the agency website chat, WhatsApp business line and the main switchboard, with separate conversation flows for candidates and clients.

The candidate flow asks for the desired role family, current role, location and right-to-work status, parses the CV when uploaded, and only escalates to a consultant when the candidate matches an active brief. Everything else is added to the talent pool with consent and routed into the relevant nurture sequence.

The client flow captures role, seniority, location, salary band, contract type and ideal start date, then offers two confirmed slots in the relevant consultant calendar. The AI never offers slots in slots already held for client meetings, and the consultant gets a one-paragraph brief in their inbox before the call.

Out-of-hours behaviour mirrors the day. The AI handles the full intake, books the call into the consultant diary for the next available window, and sends a confirmation. The morning queue is fully triaged before the team logs on.

The results

Qualified candidate and client leads tripled in the first quarter, not because traffic went up but because the AI was finally capturing and screening every enquiry instead of letting them die in inboxes.

Speed-to-lead fell from 4.2 hours to 47 seconds on average. The agency own A/B data showed that responding within a minute roughly doubled the chance of converting a first-touch into a placement conversation.

Consultants reclaimed an average of 22 hours per week per desk, which was reinvested into client development and candidate care. Placements influenced rose 38% in 90 days with the same headcount.

Cost per qualified lead dropped 61% versus the paid-social baseline, because the AI was now extracting genuine leads from existing organic and referral traffic that had previously been wasted.

Implementation playbook

The agency rolled Boafo Agent out one desk at a time, starting with the tech contract desk, which had the highest inbound volume and the most painful triage burden. The pilot ran for four weeks before any other desk was touched. That discipline is the single biggest reason the rollout went smoothly across the remaining seven desks.

Knowledge ingest was unusually fast because the agency had already invested in a structured brief template that every consultant was required to use. Importing 18 months of completed briefs gave the AI a strong baseline understanding of what a real role looks like per industry, and made the qualification rules concrete instead of abstract.

The hardest week was week three of the pilot, when the AI started escalating fewer candidates than the consultants expected. The instinct was to loosen the qualification rules, but the managing director held the line and asked the team to grade the non-escalated candidates over the following month. The grading exercise showed the AI was right in 91% of cases and the team accepted the new normal.

Consultant adoption was driven by one small UX choice: every escalated lead arrives in the consultant inbox with a single-paragraph summary at the top of the email. Consultants do not have to read the transcript unless they want to. That choice removed the most common objection to AI handoffs in professional services, which is that they create more reading rather than less.

The agency now treats the Boafo Agent dashboard as part of the weekly desk meeting. Each desk lead reviews capture rate, escalation accuracy and time-to-first-meeting, and any drift triggers a small flow change the same week. That cadence is what keeps the 3x figure intact six months after launch.

What is next

The agency is rolling Boafo Agent into the onboarding pack for every new desk it launches in 2026. The MD stated goal is that no consultant should ever again act as a switchboard for unqualified inbound traffic.

Next on the roadmap: automated reference checking and post-interview feedback collection, both currently consuming an average of three hours per consultant per week.