Back to blog

Where should a small business start with AI?

A three-step way to find where AI will genuinely help your company, before you spend on a big project.

Ever since AI became the topic of the day, most business owners have been asking the same question: “We should be doing something, but what?” The good news is that you do not need a big budget or a data science team to start. You need to ask the right question.

1. Start from repetitive work, not from the technology

List the tasks your team does the same way every week: sorting incoming emails, turning a quote into a draft invoice, preparing the monthly sales report, telling a customer where their order is. The tasks on that list that are rule-based but time-consuming are AI’s first and safest wins.

2. Run a simple check on every idea

For each task, note three things:

  • How many hours does it take each week?
  • What does a mistake cost?
  • Is the information it needs stored somewhere tidy (email, ERP, spreadsheets), or in people’s heads?

The task that takes the most hours, has low-cost mistakes and tidy data is the best candidate for a first trial.

3. Run a small trial, measure, then scale

Keep the first trial to two to four weeks, one team and one task. Agree the measure before you start, for example “cut report preparation from 4 hours to 1”. If it works, roll it out; if it does not, you lose little and learn a lot.

Do not forget data security

Pasting customer or employee data into public AI tools carries privacy and compliance risk. Writing down which tool may be used with which data should be part of the very first step.

If you are not sure where to start, our AI readiness review is designed to answer exactly that question in two weeks.

Let’s find your first step together.

In a free 30-minute call we listen to where you are and agree on the first step.