Start with the task, not the technology
The most useful question is not “how can we use AI?” but “which tasks eat our time every week?”. List the repetitive work your team does — replying to enquiries, writing product descriptions, reading invoices, preparing reports — and note how often each happens and how long it takes.
AI earns its place when a task is frequent, language-heavy and tolerant of a quick human review. It is a poor fit when every answer must be exactly right with no review, or when a simple rule or form would do the job.
Good first candidates
These tasks tend to deliver value quickly with low risk:
- Answering common customer questions from your own FAQs and policies
- Drafting replies, descriptions or summaries that a person approves
- Pulling key fields out of documents such as invoices or forms
- Sorting and routing incoming emails or enquiries
Where to be careful
Avoid fully automated decisions on pricing, legal, medical or financial matters. Keep a human in the loop, keep sensitive personal data out of prompts where you can, and choose providers whose terms say your data is not used for training.
Measure before and after
Before you start, write down how long the task takes today. After a few weeks, measure again. If the numbers do not move, simplify or stop — a useful AI project should be easy to justify.