Most small business owners we talk to are not short on AI tools. They are short on AI tools that actually get used. There is the meeting note app someone signed up for in March, the chatbot widget that answers two questions a week, and a writing assistant that three people tried once. Each one is only $30 or $50 a month, so nobody cancels them.
None of that makes AI a bad idea. It just means the buying decision usually happens backwards: a slick demo first, then a hunt for a problem it might solve. Flip it around and the decision gets much easier.
Here is the test we run with clients before anyone enters a credit card.
1. What job is this replacing, exactly?
Not "productivity". A job. "Replying to after-hours enquiries within five minutes." "Turning job notes into a quote draft." "Chasing unpaid invoices on day seven."
If you cannot name the job in one sentence, the tool will end up as another tab nobody opens. If you can, you have a yardstick for everything else on this list.
2. Who does that job today, and how long does it take?
Put a rough number on it. If your office manager spends four hours a week copying enquiries from email into the CRM, that is the prize. If the job takes ten minutes a fortnight, even a perfect tool will not move the needle.
This is also where you find out whether the real problem is the tool at all. Sometimes the four hours exist because nobody agreed on a process. AI cannot fix a process that does not exist yet. It just makes the mess faster.
3. Does it fit the systems you already run?
The best AI tool in the world is useless if it lives on an island. Ask:
- Does it connect to your email, calendar, CRM or job management system?
- Will someone have to copy and paste between it and everything else?
- What happens to your data if you cancel next year?
A lot of AI subscriptions quietly add a new manual step, which is the opposite of what you are paying for. If the answer is "we would export a CSV every Friday", be honest with yourself about whether that will really happen.
4. Where does a human check the work?
Every useful AI system we build has a clear review point. A draft quote waits for approval. A customer reply goes out only within rules you set. A listing description gets a once-over before it hits the portal.
If a tool sends things to customers or changes records with no human in the loop, ask what happens when it gets something wrong, because eventually it will. Check where your customer data goes while you are at it. Our plain-English privacy checklist covers the questions worth asking any vendor.
5. How will you know in 30 days if it worked?
Pick one measure before you start: response time, hours saved, quotes sent, invoices paid on time. Write down today's number. Put a reminder in the calendar for a month from now.
If the number has not moved, cancel it without guilt. A trial that taught you something is not wasted money. A subscription you keep paying out of habit is.
When a tool is not the answer
Sometimes all five answers point to an off-the-shelf app, and that is great. Buy it, set it up properly and move on.
Other times the answers show the job spans three systems, needs your business rules, and has to be checked by someone who knows your customers. That is usually where a small custom build beats another subscription. We wrote about the trade-offs in more detail in hiring a developer vs AI SaaS vs a local partner.
Honest constraint: no checklist replaces actually trying something. But five minutes with these questions will save you from most of the tools that end up forgotten in a drawer.
At Cloud Inn we design, build and run practical AI systems for local businesses without an IT department, on fixed fees and with human oversight by design. If you have a job in mind and are not sure whether it needs a tool, a build or neither, book a 30-minute call and we will give you a straight answer.