Use ChatGPT to clarify the problem
Begin with what your business does, who the website serves and the outcome you want. Ask for questions before solutions. An assistant can produce a convincing feature list from very little information, but a convincing list is not necessarily the right scope. Treat the conversation as a planning aid rather than a finished specification.
Provide a concrete starting prompt
Try: “I run a small court rental business. Customers currently book through messages. Help me outline a website that reduces double-bookings. Ask about availability, cancellations, staff access and payment handling before suggesting features. Mark assumptions clearly.” This gives the discussion a real problem and a boundary instead of asking for an impressive website in general.
Separate decisions from suggestions
After the questions, request three lists: confirmed requirements, proposed features and unresolved decisions. Keep them separate when sharing the brief with a developer. For example, online payment might be a suggestion while manual booking approval is a confirmed business rule. Mixing them can inflate quotations and create misunderstandings about what was agreed.
Turn the brief into acceptance checks
Ask for examples of successful and unsuccessful journeys. What happens when two customers select the same slot? What if a customer closes the browser after submitting? Review these scenarios yourself and add the operational details the assistant cannot know. A brief should explain how staff work as well as how visitors browse.
Finish with a human scope review
Remove unnecessary features and verify business facts. Use sample information rather than confidential customer records. The useful output is a document you understand and can challenge, not a long specification you feel obliged to follow. Bring that document to your developer and use it to resolve remaining assumptions.
Published by DevMania with AI assistance. Examples illustrate planning approaches and do not claim client results. Check current provider documentation before implementing product-specific features.