Part 6: Making AI do things

What AI is already built into the software you pay for?

Written lesson · about 3 minutes to read

The short answer

More than most owners realise: office suites such as Microsoft 365 and Google Workspace, and many accounting, CRM, scheduling and phone systems, now include AI features, some inside your plan and some as a paid add-on. Before you buy a new tool, list what you already pay for and check what its AI can do.

Three things to remember

  1. 01Many tools you already pay for have added AI, some included in the plan and some as an add-on.
  2. 02Built-in AI works on the data already in that system, which is its main advantage.
  3. 03Before buying anything new, check your current software's AI settings, add-on price and data terms.

What is probably already there

Two well-known examples are Microsoft 365, which has an assistant called Copilot, and Google Workspace, which has one called Gemini. Both put an AI assistant inside email, documents, spreadsheets and meetings. The same has happened across the software a small business runs: accounting, CRM (the customer records system), field service and scheduling, phone systems, website builders and ERP (the system that runs orders, stock and invoices).

Typical features are drafting a reply, summarising a customer's history, transcribing a call, or reading an invoice into an entry. Some come with your plan, some are a paid add-on, and some stay switched off until an administrator turns them on. The only way to know what you have is to look.

Why built-in AI is worth trying first

It works on the data already in that system. The assistant in your CRM can see the customer's history, and the one in your accounting software can see the invoices. There is no separate tool to buy and no files to export. Because the software maker builds it in, it is a feature you switch on, not a custom project.

The limits are real. It usually sees only its own system, so a question that spans the CRM and the accounts may need the two connected first. Quality varies from one product to the next, so test it on your own real work before you rely on it.

What to check before you switch anything on

Check four things. What does the feature actually do? What does it cost, and is that per person or per use? What are the data terms, including whether your information is used to train the provider's models? And who has to switch it on?

Then ask the vendor plainly whether an AI can read and act on your data, what it can change, and who controls that. Start with features that read and draft while a person decides. Let anything that changes records wait until the reading has proved itself.

Where to start this week

Write down every piece of software you pay for. Your card statement is a quick way to find them all. Beside each one, note whether it has AI features, what they cost and who administers the account.

Pick one feature that matches a task you repeat, such as summarising customer notes or drafting invoice reminders. Try it for two weeks with a person checking the output. If it falls short, you will at least know what a new tool would have to do better.

Next step

Where does this fit your business?

Six quick questions about your business, then your AI plan: what to start with, what a person still checks, and the lessons that go with it. Or bring it straight to a free thirty-minute call.