Sample report
What you actually receive
A worked example. Harrow & Vale Supplies does not exist, but every pattern in here does, and this is the shape a real report takes.
This is an example, not a client's report. We never publish a client's figures, redacted or otherwise. So the business below is invented, and the numbers are the ones we would typically expect at this size rather than any real company's. Download it as a PDF if you would rather read it that way.
The business
Harrow & Vale Supplies is a plumbing and heating merchant. Twenty-four staff, one trade counter, three delivery vans, roughly £4.2m turnover. It has grown steadily for nine years, and nothing about how it runs was ever designed. It accumulated.
What the office runs on today:
- A shared inbox for trade orders, plus the counter and the phone
- An Excel workbook, Orders 2026.xlsx, which is the real system of record
- Xero for invoicing and accounts
- A WhatsApp group for the drivers
- A shared drive holding delivery notes as scanned PDFs
What we looked at, and how
Four roles: the trade counter, the office coordinator, the accounts assistant, and the transport supervisor. Nine people do those four jobs. We sat with each of them for about 45 minutes while they worked, and timed a sample of real tasks rather than asking anyone to estimate.
Volumes came from Xero, not from us. Over the twelve months to March, Harrow & Vale raised 1,840 sales invoices. That is the multiplier behind every figure below, and you can check it in your own system in about a minute.
What an hour of staff time costs
We use fully-loaded cost, meaning salary plus employer's national insurance and pension. It is not a billable rate, because nobody here is billed out. For the office coordinator:
Salary £28,000 + on-costs at 19% (£5,320) = £33,320
Working year: 52 weeks − 5.6 weeks holiday and bank holidays = 46.4 weeks × 37.5 hours = 1,740 hours
£33,320 ÷ 1,740 = £19.15 per hour
If you think that figure is wrong for your business, substitute your own and the rest of the report still works. That is why it is shown.
Finding 1: the same order, typed three times
A trade order arrives by email. The coordinator reads it, types the customer and the lines into Orders 2026.xlsx, then types the same lines again into Xero to raise the invoice, then types the delivery address a third time into the WhatsApp message for the driver.
We timed fourteen orders across two days. We did not choose them: they were consecutive orders, taken as they came in.
| Step | Median | Range |
|---|---|---|
| Read the email, check the account is in credit | 2m 10s | 1m 05s – 6m 20s |
| Enter the order into the spreadsheet | 4m 25s | 2m 40s – 9m 10s |
| Re-enter the same lines into Xero | 3m 40s | 2m 15s – 7m 45s |
| Message the driver with the address | 0m 55s | 0m 30s – 2m 05s |
| Per order | 11m 10s |
We use the median, not the mean. Two of the fourteen orders were unusual: one was a 40-line refit and one a disputed account. Averaging those in would inflate the figure in a way you would be right to challenge.
11m 10s × 1,840 orders = 20,547 minutes = 342 hours a year
342 hours × £19.15 = £6,549 a year
Of that, the genuinely duplicated part is entering the same lines a second and third time. That comes to 4m 35s per order, or 141 hours and £2,700 a year. The rest is work that would still need doing in any system.
We report the smaller number as the recoverable one. A vendor would report the larger.
The other findings, in brief
Each of these was measured the same way: a timed sample, multiplied by a volume taken from the company's own records, at the fully-loaded rate of whoever does that job. The full report shows the working for each; this summary is what the first page looks like.
| What is happening | Hours / year | Cost / year |
|---|---|---|
| Re-entering order lines into Xero and WhatsApp | 141 | £2,700 |
| Chasing drivers for delivery status, because it is recorded nowhere | 164 | £3,141 |
| Matching scanned delivery notes to invoices at month end | 198 | £4,158 |
| Rebuilding the price list across three documents | 52 | £996 |
| Recoverable staff time | 555 | £10,995 |
That total is stated conservatively, and it deliberately excludes anything we could not measure: the cost of the four pricing errors found in the sample, and the fact that one person holds the whole order process in her head.
Those are real risks and they belong in the report as risks. They do not belong in the total, because we cannot put a defensible number on them.
Finding 5: software you have stopped using
Twelve months of card statements, against current user lists. This takes an afternoon and needs no implementation at all. It is recoverable next month.
| What | Why | Cost / year |
|---|---|---|
| Job management platform, 4 seats | Staff who left in 2024 and 2025 | £1,392 |
| Second cloud storage subscription | Duplicates the shared drive; last file added 14 months ago | £504 |
| Accounting plan tier | Upgraded for multi-currency, never used | £288 |
| Recoverable immediately | £2,184 |
What we recommended
First: cancel things, this week
£2,184 a year, no project required, no disruption to anyone. We say this first because it is the only recommendation that costs nothing to act on, and because a report that opens with a build proposal has told you what it wants before it has told you what it found.
Then: one order form instead of three
A single screen where an order is entered once. It creates the order record, raises the draft invoice in Xero through their API, and puts the delivery on a board the drivers can see and update from a phone. The delivery note is attached to the order, so month-end matching stops being a task.
The proposal specifies each screen, what it captures, what it writes to which system and who gets notified. There is enough detail in it for any competent developer to quote from without re-interviewing the staff.
What we considered and did not recommend
A full trade-merchant ERP would do all of this. For Harrow & Vale it would cost more in licences alone than the problem is worth, and would replace a great deal that works fine. We priced it so the comparison is in the report rather than left as an assumption.
We also looked at whether an off-the-shelf order tool could do it without a build. Two came close. Both failed on the Xero line-level sync, which is the actual problem, so they would have left the re-keying in place while adding a subscription. If either had worked, that would have been the recommendation, and this report would have been shorter and cheaper for everyone.
What it costs, and when it pays back
| Recoverable staff time | £10,995 / year |
| Software cancelled | £2,184 / year |
| Total identified | £13,179 / year |
Indicative build cost: £16,000 to £21,000, plus hosting and support of roughly £90 a month. Indicative means indicative. A firm quote follows a detailed scoping conversation, and that is a separate engagement.
On the midpoint that is a payback of about twenty months on the time alone, or a couple of months more once the hosting is paid for. The cancelled software is not counted towards it, because that saving arrives whether anything is built or not; neither are the pricing errors or the key-person risk. Whether that is worth doing is Harrow & Vale's decision, not ours, and the report says so.
Why the arithmetic is all on show
Because you should be able to disagree with it. Every figure here has its source next to it: the timed sample, the invoice count from your own system, the salary you told us. If you think the coordinator's hourly cost is wrong, change it and the report still works.
A number you cannot reproduce is a number you will not act on, and quite right too.