The situation
The business is an independent ready-mix concrete supplier in the north west, running two plants and a fleet of fourteen mixer trucks. It has been owned by the same family for thirty years and supplies a mix of house-builders, groundworkers, commercial contractors and a steady stream of one-off trade customers who need a load for a foundation or a driveway. The order book is healthy. The product is good. The relationships go back decades.
The problem was the phone. Concrete is perishable, and a builder who rings at seven in the morning asking for a load at ten needs an answer in minutes, not hours. The office had two people fielding calls, checking the day's plant schedule, working out whether a truck could be cleaned down and rerouted, confirming the mix spec, and pricing the load against whatever standing arrangement that customer had. On a busy morning the phones rang constantly, lines sat on hold, and builders who could not get through within a couple of minutes rang the next supplier in their phone.
The owner had a rough idea of the cost but had never really measured it. When we sat in the office for half a day and logged every call, the picture was worse than anyone had guessed. Roughly one in five inbound enquiries was being lost outright, either because nobody picked up in time or because the caller could not get a confirmed slot before their window closed. Another chunk were being taken down on paper, rung back later, and then cancelled because the customer had already found a load elsewhere. The business was turning away work it was perfectly placed to supply, not because the plants were full, but because the office could not get to the answer fast enough.
What we did
We did not try to take the dispatcher out of the loop. That was the first thing the owner made clear, and it was the right call. Dispatchers at a ready-mix plant hold dozens of constraints in their head at once, and the ones who have been doing it twenty years make judgement calls that no tool is going to match on day one. What we did was get them to the answer faster.
We set up an inbound capture system that handles the first touch on every enquiry, whether it comes in by phone, WhatsApp or email. It takes the order details, checks live plant capacity for the requested day and slot, checks fleet availability, pulls the customer's standing price list and mix preferences, and produces a single screen the dispatcher can look at and say yes or no in seconds. If the caller is an existing customer, the tool already knows their usual grade, their usual site, and their credit position. If it is a new customer, it captures the postcode and the spec and flags it for the dispatcher to decide.
For straightforward loads to known customers, the tool can offer a confirmed slot directly in the first conversation, with the dispatcher approving it in the background. For anything unusual, large volumes, pours with access complications, new accounts, awkward mixes, it pulls the request into a queue where the dispatcher makes the call in full knowledge of what else is on the schedule. Nothing ships without a human confirming, but the human is no longer starting from a blank sheet every time.
We also wired the system into the delivery ticket flow so that when a load is confirmed, the driver gets the job on their device with the mix, the site details, the access notes and the customer contact, and the office gets the ticket ready for invoicing the moment the wheels leave the yard.
The result
The numbers from the first three months were clean. The rate of inbound enquiries confirmed on the first call went from around forty percent to seventy eight percent. Orders lost to response time, which the business had been tracking for the first time since the project began, dropped by nearly two thirds. The two office staff who had been drowning in phones and handwritten order pads were able to shift their attention to credit control and key account work, which had been quietly neglected for years.
The recovered revenue was the headline number. Measured against the previous year's missed-enquiry rate and the average order value, the business calculated it was holding on to roughly four hundred and twenty thousand pounds of annual sales it had previously been losing to competitors on response time alone. The plants were not any busier. The fleet was not any larger. The same people were doing the same work. They were just no longer watching orders walk out the door because the phones were jammed.
The owner's take was short. "We always knew we were losing work on the phones. We just did not know how much. Now the lads in the office get to actually talk to our customers instead of apologising for not ringing back."
Frequently asked questions
Will an AI tool handle the judgement an experienced dispatcher makes?
No, and that was the brief. Dispatchers at a ready-mix plant hold dozens of constraints in their head and the experienced ones make calls no tool will match. We did not try to replace that. We removed the time they were spending hunting for information so the judgement happens faster, on a single screen, with everything they need already pulled together. Nothing ships without a human confirming.
What happens when a customer calls with an unusual order or a new account?
Anything outside the routine, large volumes, complicated access, new accounts, awkward mixes, drops into a queue for the dispatcher to decide on. The tool flags what is unusual and why, and pulls the standing schedule and fleet position alongside it. The dispatcher gets the answer in seconds rather than spending five minutes assembling it. Routine loads to known customers can be confirmed in the first call; unusual ones never are.
Does this require new phone systems or order software?
No. The capture system sits alongside however orders currently come in, whether that is the existing phone line, WhatsApp, email, or a mix. It writes back into the delivery ticket flow you already use, so the driver gets the job on their device with the mix, the site details and the access notes the same way they always did. The plants and the fleet did not change at all.
How long before fewer orders walk out the door?
The change is usually visible in the first month, because the bottleneck is response time and the new flow shortens that immediately. The bigger numbers, like the thirty-eight percent lift in first-call confirmations and the recovered annual revenue, take three months to land cleanly because you need a representative sample of busy mornings to measure against. The first sign is the office answering more calls and apologising less.