AI in business: where should an SME start?

Where to start with artificial intelligence in your business: one process, a small test, then scale. A practical three-step guide for SMEs.

Many small and medium-sized businesses know that artificial intelligence can save them time, but they don't know where to begin.

The good news is that it takes neither a large investment nor in-house technical expertise: it takes a method. The technical skills follow naturally, once the team has an excellent software stack and someone who can pull the threads of the project together. At the start, AI and the way it is used need a dedicated task force, until everyone is using it in a way that allows every process to be optimised.

Here is how we would go about it if we were advising a newcomer on using AI in their business processes, in three steps.

 

1. Start with a process

Bringing artificial intelligence into a business doesn't start with the tool, but with a process.

Before choosing any software, it pays to look at where time is lost every week: answering the same customer questions over and over, preparing near-identical quotes, sorting emails, copying data by hand from one file to another.

  • Make a short list, noting next to each item the estimated hours per month and who handles it today.

The right starting point for an SME is the most repetitive and most measurable task, not the most complex one. One well-chosen process is worth more than ten ideas left half-finished.

2. Test small

Then you test on a small scale. A business chatbot or a simple automation is trialled on that one process only, for a set period, for example four weeks.

  • Before you begin, fix a number to compare against the starting point: time saved, requests handled without intervention, errors avoided.

  • The people who work on that process should be involved from day one, because they are the ones who know where the tool gets things wrong and where it genuinely helps in everyday work. At the end of the test, the data tells you whether to continue, change course or stop, without having invested more than necessary.

3. Scale with method

If the test works, you extend it.

  • The tool is connected to the systems the business already uses, such as CRM, management software, calendar or email, and you move on to the next process on the list.

  • This is the stage where costs and priorities are really decided: how much to integrate, what to automate first, which data is needed and who manages it internally.

That is why AI consulting is most useful here, when a successful test needs to become a method that holds up over time. You grow one step at a time, with clear numbers and with the people on your team at the heart of the work.

Next
Next

Project management you barely notice: How we run projects - listen first, then structure