Not every problem needs AI. List the three things your team repeats every week and the answer is usually sitting right there.
- Date
- Reading
- 6 min read
- Written by
- Nick Li
- Tags
- #AI#Strategy
Over the past year almost every client has asked the same question: "Should we be using AI?" My answer is usually to reframe it. Don't ask whether; ask where.
Start with the repetitive work
The most direct value of AI is not replacing people. It is taking over work that is rule-based, high-volume and dull. Ask your team to list the three things they repeat every week: answering similar customer enquiries, copying email details into a system, assembling quotations. Those are almost always the best starting points.
Validate small before you invest
You do not need to rebuild your systems on day one. A support assistant embedded in your existing website, or a workflow that triages an inbox, can be live in two to three weeks. Once you have real numbers, the investment decision becomes much easier.
Get your data in order
An AI tool is only as good as what it can see. Knowledge scattered across WhatsApp threads, spreadsheets and someone's head needs to be gathered into a form a system can read. It is unglamorous work, and it sets the ceiling for the whole project.
Keep a human in the loop
Early on, let the AI draft and let a colleague approve before anything goes out. It is safe, and it builds confidence. Loosen the reins as accuracy proves itself.
A good AI project is one where, three months in, the team says: "I don't want to go back to the old way."
If you are weighing up AI for your business, talk to us. We listen first, then decide together whether it is worth doing.