It depends on how much repetitive admin you have, and nobody can tell you that without looking. So this post doesn’t give you our figure. It shows you how to work out yours.
The example below is made up. It is not a client of Fenwick Automation and the numbers are not measured results. Put your own in.
The example business
A construction or trade contractor with 20 staff.
- Two people spend part of every day on admin.
- A shared inbox with a steady stream of enquiries, orders and certificates.
- Quotes, method statements and client reports written from old versions.
- The same job details typed into the job system, a spreadsheet and the monthly report.
Step 1: add up the hours
Ask each person how long they spend on repetitive admin a day. Not all admin. Just the work that is the same every time: retyping, copying, formatting, chasing.
Say it’s 1.5 hours each.
2 people × 1.5 hours a day × 220 working days = 660 hours a year
Step 2: split it by task
Those 1.5 hours a day aren’t one job. In our example they break down like this:
| Task | Minutes a day, each person |
|---|---|
| Retyping job details between systems | 30 |
| Building documents from old versions | 25 |
| Chasing drawings, certificates and answers | 20 |
| Sorting the inbox and filing attachments | 15 |
Yours will be different. The point is to see where the time actually goes.
Step 3: mark what a system can take
Go down the list and ask one question of each task: does it need judgement, or is it the same every time?
- Retyping between systems: mostly the same every time. A system can take most of it.
- Building documents: the copying and formatting can go. The content that depends on the site or the client stays with a person.
- Chasing: a system can track what’s outstanding and draft the chaser. A person sees anything going to a client before it’s sent.
- Sorting the inbox: a system can match most emails to jobs. Anything it can’t match goes to a person.
Say that works out at half the time. In the example, that’s 330 hours a year back. That’s an assumption for the example, not a promise. The real share only shows up when you look at the actual documents.
Step 4: mark what shouldn’t be automated
Some of the time should stay exactly where it is.
- The price. The margin lives in the exceptions: access, programme, the client’s payment record.
- The safe method of work. A competent person decides it and signs.
- Anything sent to a client. A person sees it first.
We never let a system run unchecked where a mistake costs money or safety. Leave those hours out of your saving.
Step 5: turn the hours into a cost
Multiply the hours by the fully loaded hourly cost of the people doing the work. Not their salary divided by the hours in a year. If it’s a supervisor doing it at the kitchen table on a Sunday, the real cost is the site work they didn’t do.
The calculator does this with your own figures and emails you the report.
If the answer is small, it isn’t worth automating. We’ll tell you that on the call.
How Fenwick Automation would assess your business
- Find the repetitive work. On the 10-minute call, you tell me which document eats the most time. I tell you whether it fits and give you a rough estimate.
- Map the process as it actually runs. In the paid audit we look at your real documents and data, not a description of them.
- Work out the annual cost and what to build first. The audit gives you what the paperwork costs a year in your own figures, in writing.
- Check how your systems connect. We find out which of your existing software a system can read from and write to, and where a person still pastes.
- Build one route. Fixed price, from the audit.
- Test it against the real work. The parallel run: the new system runs alongside the old way until your team trusts it.
- Keep people in control. A person checks anything that matters, for as long as the system runs.
For the tasks behind these numbers, see 10 construction admin tasks you can automate. For what we are and how we work, read bespoke AI automation for UK construction and trade businesses, or see how we build AI automation and who we are.
What will go wrong
- The first estimate of daily admin time is usually a guess. Time it for a week.
- Two people doing the same task will do it differently.
- The saving is smaller if your data is scattered, because the first weeks go on tidying it.
Next step
Book the 10-minute call. Bring the number from Step 1. You’ll leave knowing whether it’s worth building and a rough estimate of what it would cost.