This is an illustrative example, built from the kind of work we do rather than one specific named client. It shows how our method plays out in practice, with realistic numbers, on a business that will feel familiar to a lot of owners.
Picture a land and measured surveying firm we worked with. Surveyors out most days with drones, scanners and total stations, capturing sites and turning them into reports. 22 staff, 14 of them surveyors. Annual revenue of £2.4M. On paper, a solid business. Underneath, a business where the founder had become the bottleneck without quite realising it.
The starting point
The firm was getting 40 enquiries a month and winning 28% of the quotes it sent out. Only 60% of surveyor time was billable, the rest lost to travel, admin and waiting on approvals. Finished jobs were taking 12 weeks from completion to being paid, and at any given moment there was £95,000 of finished work sitting unbilled. That's real money, earned, just not collected.
The reason wasn't a lack of work ethic. It was that the founder priced and signed off every single quote himself. One quote, for a job that should have been straightforward, took 3 days to go out. A competitor got theirs in faster and won the job. Nothing wrong with the surveying. Everything held up in the office.
Step one: the AI review
We started with an AI review, half a day spent mapping where the biggest leaks actually were. Not a guess. A proper look at quoting, billing, supplier spend and how surveyor time was actually being used. That map pointed straight at two things: money leaking out through supplier spend, and time leaking out through a quoting process that ran through one person.
Step two: shared savings
Before touching any process or any tool, we went through 12 months of spend. The firm was spending £410,000 a year across its suppliers. We found £41,000 a year in savings, spread across the categories that make up most surveying overheads:
- Vans, leases and fuel cards: spend £120,000, saving £13,000
- Subcontract surveyors: spend £90,000, saving £8,500
- Equipment hire and calibration: spend £70,000, saving £7,500
- Indemnity and insurance: spend £55,000, saving £5,000
- CAD, GIS and software: spend £45,000, saving £5,500
- Everything else: spend £30,000, saving £1,500
£41,000 a year, found before anything else was built or changed. That's the point of doing this step first. The money to fund the rest of the project was already sitting inside the business, just going out the door unnoticed.
Step three: getting the process out of the founder's head
Next came the part that doesn't involve any software at all. We sat down with the founder and went through 10 past quotes, one at a time, asking why he'd priced each one the way he had. Site access, ground conditions, equipment needed, turnaround time, risk. All of it was in his head. None of it was written down anywhere.
We turned that into a set of pricing rules on paper, then tested them against live enquiries coming in, adjusting where the rules didn't quite match reality. We didn't move to the next step until those rules held up consistently.
Quoting no longer depended on one person being available, in the right mood, with the right job in front of them at the right time.
Step four: the tool, built with their own team
With the rules written and tested, we built an AI quoting tool alongside someone on the firm's own team. The tool drafts quotes against the written rules. A person still checks and sends every one, nothing goes to a customer untouched. The tool is wired into the scheduling board too, so when a report goes out, the invoice is raised the same day. No one has to remember to bill the job. It happens because the job's marked done.
The result after twelve months
Twelve months on, the numbers had moved across the board:
- Time to send a quote: from 3 days to same day
- Quotes won: from 28% to 35%
- Surveyor time billable: from 60% to 72%
- Finished work not yet invoiced: from £95,000 to under £45,000
- Finished job to money in the bank: from 12 weeks to 10 weeks
- Supplier costs: from £410,000 to £369,000
At the same 40 enquiries a month, a rise from 28% to 35% works out at roughly three extra jobs won a month, with nothing extra spent on marketing. At an average job value of around £4,500, that's roughly £160,000 a year in extra work won, simply because quotes went out faster and more consistently.
How the numbers paid for themselves
The £41,000 a year in supplier savings didn't just sit there looking good on a spreadsheet. It funded the rest of the project. The shared savings fee came to £20,500. The process work and tool build came to £9,500. That left £11,000 in the business from savings alone in year one, after every fee was paid.
From year two onwards, there are no further fees against that saving. The full £41,000 a year stays with the business. Over three years, once all fees are accounted for, that's £93,000 kept from supplier savings alone. On top of that, around £50,000 in cash was freed up from work that had been finished but sat unbilled, plus the extra work won from the higher quote win rate. These figures are illustrative, built to show how the pieces fit together. But the shape holds in real cases: the savings pay for the project, and what's left over is money that wasn't there before.
Why the order mattered here
None of this started with buying an AI tool. It started with finding money already being wasted, then getting the founder's pricing knowledge onto paper, then building something to run those rules with a person still checking every quote. Build the tool first, on top of a process that lived only in someone's head, and you just automate the same bottleneck. Do the groundwork first, and the tool actually earns its place.
If any of this sounds familiar, a founder who prices everything, quotes that take too long to go out, jobs finished but not billed, it's worth finding out what's happening in your own numbers before spending anything on a tool.
Talk to us and see what's actually costing you.