There is a lot of pressure on small business owners right now to "do something with AI".
Most of that pressure comes from people selling AI.
Here is the honest version.
Automating a mess gets you a faster mess
If your quoting process is inconsistent, an AI that produces quotes will produce inconsistent quotes at speed.
If nobody agrees on what a job costs, a tool that calculates job costs will confidently produce a number nobody trusts.
The technology does not supply the judgement. It supplies the throughput. So the sequence matters:
- Work out what the process should be. Written down, agreed, boring.
- Simplify it. Most processes have steps that exist because someone left in 2019.
- Then automate what is left.
Skip step one and you buy an expensive way to make the same mistake more often.
What is actually worth automating
Not the interesting stuff. The repetitive stuff that eats your week.
- Quoting. Same inputs, same maths, every time. This one usually pays for itself first.
- Scheduling and reminders. The follow-up nobody has time to make.
- Invoicing and chasing. Especially the chasing, which owners hate doing and therefore do late.
- Getting information out of documents. Supplier invoices, timesheets, site reports.
- First-draft writing. Job descriptions, policies, standard emails. A draft to react to beats a blank page.
Notice what is not on that list. Anything requiring judgement about your customers, your people, or your pricing.
A test worth applying
Before automating anything, ask: could I hand this task to a new employee with a one-page instruction sheet?
If yes, it is a good automation candidate. The process is clear enough to be handed over.
If no, the process is not ready. Fix that first, and you will get value out of it whether you automate it or not.
What this looks like in practice
Most of the businesses we work with are not building anything exotic.
They are getting onto one platform instead of five that do not talk to each other. They are removing the double entry. They are putting a tool in place so the quote that used to take Saturday morning takes fifteen minutes on Thursday.
That is not a headline. It is a weekend back, and a business that stops losing money in the gaps between systems.
Start there. The clever stuff is still available later, and it works far better on top of a process that already makes sense.
