Are you still paying for software your AI can now do?

The quickest way to find out is the one-sentence test. Pull up everything your business pays for each month and try to explain what each subscription does in one sentence. The ones that make you hesitate, or that you cannot explain at all, are the ones to look at properly. Do not cancel anything yet. The first job is to find out which tools you could not defend.

Why look at your software stack now?

Because a lot of stacks were assembled before AI could do any of the work. There was a tool for finding prospects, a tool for tidying the list, a tool for writing the first message, and a tool for scheduling it. Each was a sensible purchase at the time.

Very few have been looked at since an assistant the business already pays for became able to do part of that job. Subscriptions are easy to start and easy to forget, so they tend to accumulate. Nothing is wrong with any single one. The question is whether the whole set still makes sense.

What is the one-sentence test?

It is as simple as it sounds.

List every subscription across sales, marketing and operations. Next to each, write one sentence for what it does for the business. Not what it can do, what it actually does here.

Some will be easy: "it sends the invoices". Some will make you pause. A few you will not be able to explain at all, because somebody set them up two years ago and they have been quietly renewing ever since.

What do the hesitations tell you?

Mark every tool where the honest sentence was "we have always had it" or "someone set it up". Those go on a short list to look at properly.

For each one on the list, ask three things. Is anyone using it this month? Does it do something an assistant you already pay for could now do well enough? And if you switched it off for a month, who would notice?

If the answers are no, maybe and nobody, you have found something worth a closer look.

Should I just cancel everything AI can replace?

No. Some tools do a job AI should not, such as anything holding records you must keep accurate or anything that has to run the same way every time. For work like that, a plain workflow beats an assistant, and the choice between a workflow and an agent is worth understanding before you swap anything.

It is also worth being wary of anyone who tells you your stack is bloated, or that it is fine, especially a software vendor writing about their own product. Comparisons online are often written to sell something. The point of the test is to find out for yourself, on your own business, with your own sentences.

What about the cost of switching?

Real, and worth counting. A tool that costs a little but saves a person an hour a week is cheap. A tool that costs a lot and nobody opens is not. Switching also takes time and carries risk, so we prefer to change one thing at a time and check it works before removing the old one.

This is why we say do not cancel anything in the first week. The goal of week one is knowledge, not savings. You will often find the quick win is not a cancellation at all, but a tool you pay for and use at ten per cent of what it can do.

How do I turn the list into a plan?

Rank the short list by cost and by how easy it would be to replace, then pick one. Try the replacement alongside the old tool for a month. If it holds up, switch. If not, you have lost nothing and learned what the tool was really for.

At Handiwork, when we look at a business's tools, the one-sentence test is the first thing we do. It takes about an hour and it makes every later decision easier. If you are not sure where AI fits at all, where to start with AI is a good first read, and our use cases show what other small businesses have done.

What to do this week

Make the list, write the sentence, and mark the ones where you hesitated. Do not cancel anything this week. Just find out which ones you could not defend.

Ready to find out where you stand?

If you want an outside read on which of your tools earn their place, our free AI Readiness Check takes about five minutes: AI Readiness Check.