LESSON 05

HOW TO JUDGE A TOOL
WITHOUT BEING TECHNICAL.

You do not need to evaluate the technology. You need to evaluate the business decision, and that is something you already know how to do in every other category of purchase.

Learn AI › Choosing Tools

Start with the problem, not the demo

The demo is designed to be impressive. It was built by people whose job was to make it impressive, using data chosen to make it look good. Watching one tells you almost nothing about whether it will work in your business.

Before a single demo, write one sentence: the specific problem this would solve is _____, and I know it is a problem because _____. If you cannot complete that sentence with something you have actually observed this month, you are shopping, not solving.

The second half matters as much as the first. 'Our follow-up is inconsistent' is a belief. 'Eleven leads from last month were never contacted a second time' is an observation. Only the second one lets you tell later whether the tool worked.

The five questions

  1. What specific problem does this solve, and is it one of my top three? Most tools solve a real problem that is seventh on your list. Solving your seventh problem costs the same attention as solving your first.
  2. Whose data goes into it, and where does it live? Customer information, seller information, financial detail. Can you get it back out. What happens to it if you cancel.
  3. What happens when it is wrong? Not if. Who catches it, how fast, and what does the mistake cost between happening and being caught.
  4. What does this cost at ten times the volume? Per-seat and per-use pricing behave very differently as you grow, and the one that is cheap today is often the one that is punishing at scale.
  5. How hard is it to leave? Ask this before you sign, because you will never have more leverage or less emotional investment than you do now.

Nothing on that list is technical. Every one of them is a question you would ask about a vendor, a lender, or a subcontractor.

The exit question, in more detail

This is the one people skip and the one that hurts.

Switching costs accumulate quietly. Your data ends up in a format only that tool reads. Your team learns an interface. Your process bends around how the software wants things done. A year in, the tool is mediocre and leaving it means a month of disruption, so you stay, and you keep paying for mediocre.

Ask three things before you commit. Can I export my data in a format something else can read, and can I test that this week rather than take your word for it. What is the contract term and what happens at renewal. If I cancel, how long do I have access to my own records.

A vendor confident in their product answers all three without friction. Hesitation on the export question in particular is worth treating as a finding.

Signals that a tool is worth your time

Good signalBad signal
They ask about your process before pitchingThey lead with a feature list
They can name who this is not forIt is for every business in every industry
Clear, published pricingPricing available after a call
A trial you can run on your own dataA guided demo only
Straight answers about data handlingVague reassurance about security
Support that includes a personSupport that is a help center and a form

The pattern is the same one that applies to buying anything. A seller who narrows the fit is telling you the truth about the product. A seller who says it works for everyone is telling you about their sales targets.

Buy less than you think you need

The most common mistake is not buying the wrong tool. It is buying four tools in a quarter.

Every tool carries a cost that never appears on the invoice. Someone has to learn it, configure it, maintain it, and remember it exists. Four tools half-adopted produce less value than one tool used properly, and they produce considerably more confusion, because now nobody is sure which system holds the real answer.

A reasonable pace for a small business is one new tool at a time, adopted fully, measured against the problem you wrote down, before the next one is considered. Slower than it feels like it should be. Faster in the end, because you are not carrying four partial implementations.

And before any of it: if the process the tool would run is not written down, the tool is premature. Software executes a process. It does not supply one.

Frequently asked

Questions people actually ask

Should I buy an AI tool or use a general assistant?

Start with a general assistant, because it costs almost nothing and teaches you what you actually need. Purpose-built tools earn their place when you have a specific, repeated, well-defined workflow that a general tool handles clumsily.

How do I evaluate a tool if I cannot judge the technology?

You are not judging the technology. You are judging a vendor, a fit to a problem you have already named, a cost curve, and an exit. Those are ordinary business judgments and you already make them.

Is free AI software a bad idea for business use?

Not automatically, but read what the free tier does with your data, because that is usually where the price is. For anything touching customer or financial information, a paid business plan with clear data terms is the cheaper decision.

What if my team will not use it?

Then it does not matter how good it is. Adoption depends on whether it removes work someone actually finds tedious. A tool that adds a step so a dashboard can be complete will be abandoned by week three, correctly.

How many tools should a small business run?

Fewer than it has. The practical limit is how many one person can keep configured and current. For most small teams that is a very small number, and the discipline of staying under it is worth more than any individual tool.

Make your next move

A year from now, what will you be glad you started today?

You don't need another promise that everything will be easy. You need something useful to learn — and a next step you're willing to take.