Is AI automation worth it for a small business?

Sometimes. The honest version of the answer is that AI automation pays for itself in a small business when it removes a particular kind of job, and wastes money when it is bought as a thing to have. The difference is easy to see in advance if you look at the work rather than the technology.

The test

A job is worth automating when it is three things at once.

Repeated. It happens every week or every month, in roughly the same shape. A one-off, however painful, is cheaper to do by hand.

Dated. It has to happen by a time: the reminder before the deadline, the report on Monday, the check before the menu goes to print. Dated work is what people forget and what software never does.

Documentable. Someone can write down how it is done, including the exceptions. If the job lives entirely in one person's judgement, automation will either get it wrong or push the judgement back onto them.

Chasing finalists for their photos and bios across nine awards programmes passes all three. Watching what competing pubs have done to their menus passes all three; it is one of the tools The Plough in Harborne runs on a schedule. Deciding whether to take on a new client passes none of them, and should not be automated.

Where it pays

A few patterns come up again and again.

• Chasing. Reminders, confirmations and follow-ups that someone currently sends by hand from a list. Drafted in your voice from your records and sent on a schedule, ideally with a person approving.
• Watching. Competitors' prices, a regulator's updates, a supplier's website, a deadline register. A monitor on a schedule that tells you what changed replaces a job nobody has time for.
• Drafting. Letters, reports, updates and summaries written from data you already hold, for a person to check and send. The time saved is in the first draft, not the final word.
• Reporting. The weekly spreadsheet somebody assembles from three systems. Built once, it builds itself.

In each case the saving is measured in hours a week, and the hours belong to someone whose time has an obvious value.

Where it does not

• Point solutions bought for their own sake. A chatbot on the website, a transcription tool nobody reads, an "AI assistant" that answers questions nobody was asking. These can impress in a demo and then go unused.
• Jobs with no record. If the process exists only in someone's head, automate the record-keeping first and the job second.
• Anything client-facing without review. An automated message that goes wrong goes wrong at scale. Everything that leaves the building should pass a person first.
• Replacing the judgement. The admin around a decision is automatable. The decision usually is not, and pretending otherwise is how businesses end up with a system nobody trusts.

How to start without buying anything

Begin with an audit, not a tool. Half a day with whoever does the work, listing every repeated job, who does it, how long it takes and what happens when it is late. Put an hours-per-week figure beside each one. The list ranks itself.

Then build the top one, small, and measure it. If the hours come down, build the second. If they do not, you have spent little and learned something true about where the time goes, which is worth having anyway.

What good looks like a year in

The business runs the same way, with the same people, and the chasing, watching, drafting and reporting happen without anyone remembering to do them. Nobody talks about AI. The system has a name, someone owns it, and the hours it gave back are being spent on the work that needed a person all along.

That is the test for whether it was worth it, and it is a number, not a feeling.

Where to start

List the repeated jobs, put hours beside each, and look at the top three. If none of them is repeated, dated and documentable, the honest answer for now is no, and that is worth knowing too.