Many businesses want to "use AI" but are unsure where to begin. A good first project is small, repetitive and low risk, with a person still making the final decision. Sorting new customer enquiries fits that description well.
The problem
Enquiries arrive through a website form or inbox. Someone reads each one, decides what it is about, and forwards it to the right person. It takes a few minutes each time, and urgent requests can wait behind routine ones.
The workflow
- A new enquiry arrives through your form or a shared inbox.
- An automation tool sends the message to an AI model with a short instruction, for example: "Classify this enquiry as Sales, Support, Billing or Other, rate its urgency as High or Normal, and write a one-sentence summary."
- The result is added to a shared sheet or task board with the original message, the category, the urgency and the summary.
- The right person is notified, and high-urgency enquiries are flagged.
- A person reviews and replies. The AI sorts; it does not answer the customer.
Why this is a good starting point
- It saves time on every single enquiry.
- Mistakes are easy to spot and correct, because a person still reads every message.
- It teaches your team how AI behaves before you use it anywhere more sensitive.
Before you build it
Decide what customer information may be sent to an AI service, and check the provider's data terms. Keep the instruction short and specific, test it on twenty real past enquiries, and adjust the categories until the results are reliable.
Once this is running smoothly, you can extend it: suggested reply drafts for common questions, or a weekly summary of what customers are asking about. Interested in learning to build workflows like this? Our AI Automation Specialist Training covers it in depth.
