29 checks across 7 systems
What we actually check.
The Leak Survey Method is a fixed procedure for costing operational losses in a small business. It rates every finding on a four‑level condition scale, states the evidence each figure rests on, and records both the things checked and found sound and the things that could not be measured at all.
Where this comes from
None of it is invented. The condition ratings follow the convention used in RICS building surveys; grading a survey by how deeply it measures comes from ASHRAE energy audit levels; and the billed‑versus‑non‑revenue split is the IWA standard water balance, borrowed from utilities that have been costing leaks for a century.
If you already know all twenty‑nine of these numbers for your own business, you don’t need us.
Water utilities split leaks into two kinds: the ones that surface, and the ones that don’t. Surfacing leaks you already know about — the no-shows, the staff standing idle. Non-surfacing leaks never announce themselves: the quote accepted in March that nobody ever booked, the customer list that quietly stopped being worked, the rejected claim that timed out in an inbox. The second kind is why a survey is worth paying for.
Enquiries & first contact5 checks
- Calls and messages missed during working hours
- Average time to a first reply, and who replies
- Enquiries that never got a second attempt
- Messages sitting unread past a day
- Walk-ins or callers nobody recorded anywhere
Capacity & scheduling5 checks
- Utilisation by hour and by weekday
- No-show and late-cancellation rate, by type of booking
- Gaps created by cancellations nobody refilled
- Days between an enquiry and the first slot offered
- Quiet periods nobody is deliberately filling
Quotes & conversion4 checks
- Quoted versus accepted
- Accepted versus actually scheduled
- High-value quotes with no follow-up owner
- Prices given verbally and never recorded
Follow-up & retention4 checks
- Customers due for contact and not contacted
- The date the list was last actually worked
- Customers lapsed beyond twelve months
- Post-delivery check-ins that never happen
Money & billing4 checks
- Work delivered but never invoiced
- Payments pending beyond sixty days
- Discounts given without a record
- Daily collection measured against daily production
Stock & supplies3 checks
- Reordering done on memory rather than on usage
- Write-offs over the last two quarters
- Emergency purchases made at premium
Data & visibility4 checks
- Numbers that live in three places and disagree
- Reports somebody rebuilds by hand every month
- Anything that exists only in one person’s head
- Decisions currently being made on gut feel
Working with a café or a commerce business? The structure is identical — the line items change to ticket times, daypart revenue, stock variance, order-to-dispatch lag and returns cost.
10 working days · typically 12–20 findings
The 10‑Day Leak Report.
Ten working days inside the business. Every finding carries a condition rating, the evidence it rests on, and a monthly figure. The same method runs on a clinic, a café or a warehouse; only the line items change.
-
Step 01Days 1–3
Observe
A full working day watched, first contact to closing. We read the books, ask questions, and change nothing.
-
Step 02Days 4–6
Trace
One customer followed end to end, and one rupee. This is where a leak stops being a feeling and gets a location.
-
Step 03Days 7–8
Quantify
Each leak gets a monthly figure from your own numbers. If we cannot price it, it does not go in.
-
Step 04Days 9–10
Report
Findings ranked by what they cost, each with its fix. Walked through in person.
What you receive
A written report listing every leak, its monthly cost, and what closing it takes. Yours to keep.
What it costs
A fixed fee, agreed in writing before we start. No retainer, no day rates.
You get the number on the 20‑minute call.
Go ahead with the build and the survey fee comes off it.
What happens after
Most owners want the top two closed. We build the smallest system that closes the largest leak — not a platform, a fix — then measure the same number again. If it did not move, we say so.
A report you can’t act on is just a bill.
Finding the leak is the easy half. Most consultants stop there, hand you a PDF and invoice you.
We build the fix. Usually small and unglamorous — a follow-up list that runs itself, a queue that clears daily. It lives inside the tools your staff already use, because a system nobody opens is another leak.
Then we measure the same number again. If it didn’t move, we say so and go back to the map.
You own the code. Some of it uses language models where that is genuinely the cheapest way to do a job; most of it does not, and the report says which. We are not selling you AI. We are selling you a number and the thing that moves it.
