How a logistics company cut missed calls by 78% with a 24/7 AI receptionist

The problem

The dispatch team was answering more than 400 inbound calls per week across three depots. Roughly 60% of those calls were repetitive: driver check-ins, ETA updates, proof-of-delivery requests and quote enquiries, yet they consumed the same dispatcher who needed to be planning the next route.

After-hours and weekend coverage was the bigger pain. The phone line rolled to voicemail between 18:00 and 07:00 and most messages were never returned within 24 hours. Larger shippers had started moving urgent loads to competitors who picked up on the first ring.

The operations director estimated that missed calls were costing the business GBP 18,000 a month in lost spot-market freight, not counting the goodwill cost of customers who churned silently after being unable to reach a human.

Hiring a traditional answering service had been tried and abandoned twice. The agents did not know the difference between a pallet and a TEU, could not quote even a rough rate, and frustrated customers who expected industry fluency. The team needed something that sounded like one of them, was available the moment a call landed, and could push qualified work straight into the dispatch board.

The solution

Boafo Agent was deployed across the company website chat widget, the main inbound phone line and a dedicated WhatsApp number. The AI receptionist was trained on the operator lane card, current spot pricing, equipment list and SOPs for common request types.

Three conversation flows were configured: an ETA self-service flow that looks up the load number and reads back live tracking, a quote intake flow that captures origin, destination, weight, equipment and ready date, and a careers flow for driver applications. Every flow ends with a structured payload pushed into the dispatch board and a confirmation message to the customer.

After-hours behaviour was made explicit. Between 18:00 and 07:00 the AI tells callers it is the on-call assistant, captures the request in full, and texts the on-call dispatcher only when the load is flagged urgent or above a value threshold. Everything else is queued for the morning with full context, so the day shift starts already triaged.

Boafo Agent also handles handover. When a caller asks for a human, the AI summarises the conversation in two sentences and warm-transfers to the next available dispatcher, so nobody starts a call cold. The operations team measured a 40% drop in average handle time on transferred calls.

The results

Within the first 30 days, missed calls fell from an average of 92 per week to 20 per week, a 78% reduction. Every after-hours enquiry was captured with a structured payload instead of a voicemail.

Quote-to-booking conversion improved by 42% in the first quarter because every web and WhatsApp enquiry now received an instant, qualified response and a callback window, instead of waiting for a dispatcher to circle back.

Dispatchers reclaimed an average of 31 hours per week across the three depots. That capacity was reinvested into live operations and proactive customer updates, both of which lifted CSAT to 4.8 out of 5.

The operator now uses Boafo Agent analytics to spot lane-level demand patterns, including a recurring Sunday-night spike in same-day requests that had been invisible when calls were going to voicemail.

Implementation playbook

The implementation playbook the operator followed is worth restating, because it has been repeated almost identically at four other logistics customers since. Week one was discovery: a single 90-minute working session with the operations director and the senior dispatcher to map every recurring call type, document the rules for what counts as urgent, and agree the escalation thresholds. No software was touched in week one. The output was a one-page intent document that everyone signed off.

Week two was knowledge ingest. The lane card, current spot pricing, equipment list, driver rota and the last 90 days of quote correspondence were imported into the Boafo Agent knowledge base. The dispatcher who ran the import flagged 14 inconsistencies in the existing pricing sheet that nobody had noticed for months, which the operations team fixed before the AI ever spoke to a customer.

Week three was shadow mode. Boafo Agent ran alongside the human team on web chat and WhatsApp only, with every draft response reviewed by a dispatcher before sending. Three days of shadow mode produced 47 small tone-of-voice corrections and two genuine knowledge gaps, both fixed the same day. The voice line went live on day 16 and the after-hours flows on day 18.

What made the deployment stick is the weekly review cadence. Every Friday the operations director spends 20 minutes in the Boafo Agent dashboard reviewing the top 10 conversations by escalation rate and the top 10 by customer sentiment. Any pattern that appears twice is addressed the same week, either with a knowledge-base update or a small flow change. That cadence is the single biggest reason the AI has kept improving instead of going stale.

Cost framing matters too. The operator did the maths upfront and committed to the deployment on a 90-day payback window. The actual payback came in 38 days, driven primarily by recovered after-hours quote revenue and the retirement of a part-time overflow contract. The board signed off year-two expansion to carrier onboarding on the back of that number.

What is next

The operations team treats Boafo Agent as a permanent member of the dispatch floor. Every new lane, equipment change or rate update is added to the AI knowledge base the same day it goes live with drivers, so the agent never goes stale.

The next phase is to extend the same conversation engine to carrier onboarding and driver check-calls, two more sources of repetitive inbound traffic that pull dispatchers off the planning board.