Most automation aimed at trades businesses is sold on the wrong thing. It gets demonstrated doing something impressive, everyone agrees it is clever, and nobody opens it again after the second week.
The version that survives is dull. It handles the job you do forty times a month that has no thinking in it, and it does that job the same way every time whether or not anyone remembers.
Here is where the hours actually are, in rough order of how much time they give back.
1. Answering the same five questions
Every trade has them. Do you service my suburb. Are you licensed and insured. How soon can someone come out. Do you charge for quotes. What does a call out cost.
These arrive by phone, by form, by text and through social media, and each one interrupts whatever was happening. The answers never change. Nothing about them requires the person who knows the trade.
Handled properly, this is not a chatbot that argues with people. It is something trained on your actual service area, your actual pricing policy and your actual availability, that answers the predictable part instantly and hands over anything it has not been told about.
The saving is not only the minutes. It is the interruption. A question answered at nine at night is a job you did not lose to whoever answered first.
2. Getting the enquiry out of the inbox
Most trades businesses have a version of the same problem. An enquiry arrives, somebody reads it on a phone while standing in a roof cavity, intends to deal with it later, and later does not happen.
The fix is unglamorous. The enquiry gets read on arrival, the details get pulled out, and it lands somewhere that is not an inbox: a job list, a spreadsheet, a CRM, whatever you already use. Nobody has to retype an address.
Routing matters as much as capture. An emergency call out and a request for a kitchen quote should not sit in the same undifferentiated pile, and deciding which is which is exactly the kind of judgement that can be automated reliably because the words people use are so consistent.
3. Following up on quotes
This is the one that pays for the rest. Most trades businesses quote well and follow up badly, and the follow up is where the margin sits.
A quote sent on Tuesday and never mentioned again is not a lost job, it is a job you stopped competing for. The customer did not decide against you. They got busy, and somebody else rang them.
Automated follow up does not mean pestering. It means the second contact actually happens, at a sensible interval, without depending on anyone having a good week. We cover this properly in its own article, because doing it badly is worse than not doing it.
- A quote that has gone quiet gets one nudge rather than none
- The nudge happens on a schedule rather than when someone remembers
- Anyone who replies goes straight to a person
- Anything already won or lost drops out of the sequence
4. The reporting nobody has time to build
Where did this month’s jobs come from. Which suburbs are worth advertising in. What proportion of quotes turned into work.
Most trades businesses cannot answer these, not because the data is missing but because assembling it takes an evening nobody has. A report that builds itself each month is a small piece of automation with a disproportionate effect, because it changes what you spend money on.
What is not worth automating
Being honest about this matters more than the list above, because the failed attempts usually come from automating the wrong things.
- Anything involving a price on a non-standard job, because getting that wrong costs more than the time it saves
- The first conversation on a large or unusual project, where the value is in a person asking the questions
- Complaints, which need a human immediately and make everything worse when they do not get one
- Anything you only do a handful of times a year, where the setup outlasts the benefit
The test is simple. If the task is frequent, predictable, and the answer does not change based on judgement, it is a candidate. If any of those three is missing, leave it alone.
Start with one
The most common failure is trying to automate everything at once, which produces a system nobody trusts and everybody works around.
Pick the single task that annoys you most this week. Automate that one thing properly, watch it for a month, and only then add the next. A business running one reliable automation is in a far better position than one running six that everyone has quietly stopped believing.
If you want a look at which of these would actually pay off in your business, that is what our AI and automation work starts with, and it usually finds two or three obvious ones in the first conversation.
