Turn on the AI you already pay for

Most small businesses already pay for AI. They just have not switched it on.

Over the last few years, nearly every kind of business software has added AI features: email, documents, spreadsheets, accounting, scheduling, customer records, phone systems, help desks. Many of those features arrived in an update nobody read about. Some are included in the plan you already pay for. Some sit behind a setting an admin has to turn on. Either way, they are easy to miss.

This post is a simple way to find them, test them and decide whether they are enough before you spend money on anything new.

Why this is the first place to look

A feature inside a tool you already use has three things going for it. It is already paid for, or close to it. It already has your data, so there is nothing to connect or export. And your team already knows where to find it, because it lives in a screen they open every day.

Compare that with a new tool or a custom build. Both mean new cost, new logins, new training and a new place for information to live. Sometimes that is worth it. But you should know what your current tools can do before you decide it is.

When we run an audit, checking current tools is one of the first steps for this reason. Sometimes the right recommendation is not a build at all. It is "turn on the feature you already have and show the team how to use it."

Step one: list what you pay for

Start with the bills. Go through the last two or three months of card and bank statements and write down every software subscription. Most businesses find more than they expected, including a few nobody uses anymore.

For each one, note three things: what it does, who uses it, and which plan you are on. The plan matters, because AI features are often included in some plans and not others.

Step two: check each tool for AI features

For each tool on the list, spend ten minutes looking. Good places to check:

• The admin or settings page, often under a heading like "AI," "assistant," "smart features" or "labs."
• The vendor's recent release notes or product updates page.
• Your plan details, to see which features are included and which cost extra.
• A quick question to the vendor's support team: "What AI features are included in our current plan, and how do we turn them on?"

Write down what you find in plain words. "Can draft email replies." "Can summarize a meeting recording." "Can read a receipt and fill in the expense." You are looking for what the feature does, not what it is called.

Step three: match features to real work

A feature is only worth turning on if it touches work your team actually does, and does often. This is where most businesses skip a step and end up with a setting switched on that nobody uses.

Take your list of features and hold it up against the tasks that eat your team's week. If someone spends hours each week writing similar replies to customers, a drafting feature in your email or help desk tool is worth a look. If someone retypes numbers from invoices into your accounting software, a feature that reads documents is worth a look. If nobody in the office writes long documents, a document-writing assistant probably is not.

If you do not know where the hours go, time the work first. A week of simple tracking will tell you more than any feature list. We wrote a separate guide on how to do that in one week.

Step four: test it on real work for two weeks

Pick one feature and one person who does the task every day. Turn the feature on for them, and ask them to use it on their real work for two weeks.

Before they start, write down roughly how long the task takes now. At the end, ask three questions:

• Did it save time, and roughly how much?
• How often did you have to fix what it produced?
• Would you be annoyed if we turned it off?

The last question is the honest one. If the answer is "not really," the feature is not earning its place, however impressive it looked in the demo.

Read the vendor's documentation on how the feature handles your data before you test it, and decide for yourself whether that fits how your business works. That is a judgment for you and your advisors, and it is worth making on purpose rather than by default.

Where built-in features fall short

Built-in AI is good at work that happens inside one tool. Drafting a reply inside your inbox. Summarizing a document inside your document editor. Suggesting a category inside your accounting software.

It tends to struggle with work that crosses tools. If a task means reading an email, looking something up in your customer records, checking a spreadsheet and then writing a document, no single vendor's feature sees the whole job. That cross-tool work is often where the biggest hours are hiding, and it is usually where a custom build earns its cost.

Built-in features also follow the vendor's idea of how the work should go. If your process is a little different, which most are, the feature may get you most of the way and then leave someone to finish by hand.

What to do with what you find

At the end of this exercise, you will have one of three results for each task you looked at:

• Solved. A feature you already pay for handles it well. Turn it on for everyone who does that task and show them how to use it.
• Partly solved. A feature helps but leaves a gap. Note the gap and how much time it still costs.
• Not solved. Nothing you own touches it. This goes on the list of candidates for a build.

The second and third groups are where a measured look pays off. Our audit maps three to five workflows, times them, and prices each fix, including whether the right answer is a feature you already own. It is $2,500 fixed and takes 5 business days.

Start with what you have

If you want a quick read on where you stand, take the free AI readiness quiz. It asks ten multiple-choice questions about your workflows, data and tools, and it will show you where the hours are most likely hiding. Or book thirty minutes and we will talk through what you already own.