The method
Use this prompt in ChatGPT or other LLMs to simulate an expert content architect. Provide the LLM with context about your project and content needs. Refine prompts based on the LLM's responses to achieve optimal content models for your Headless CMS.
The prompts
Prompt 1
You are a seasoned content architect specializing in headless CMS implementations. A client is building a new e-commerce platform for artisanal cheeses. They want to use a headless CMS to manage their product information, blog posts about cheese pairings, and customer testimonials. Considering this context, define a content model that outlines the content types, fields, and relationships needed. Include considerations for SEO, accessibility, and future scalability. The content model should be easily implementable in a modern headless CMS like Contentful, Sanity, or Strapi.
Prompt 2
Act as a content modeling expert. I'm creating a website for a non-profit organization focused on wildlife conservation. We need a headless CMS to manage information about endangered species, conservation projects, donation campaigns, and volunteer opportunities. Suggest a detailed content model with specific fields and relationships between content types. Consider the different ways users will interact with the content (e.g., searching for animals, donating to a project, signing up for volunteer events). Justify your content model choices and explain how they will support the organization's goals.
Prompt 3
Assume the role of a lead architect responsible for crafting content strategy and architecture. My organisation is launching a new educational platform that provides courses on various technical topics. We have selected Contentstack as our Headless CMS. Design a robust, structured content model for managing course information, lessons, quizzes, and instructor profiles. Ensure the design facilitates content reuse, localization, and integration with other systems. Include considerations for version control, user roles, and workflows. Specify what fields should be modular content blocks. How should we organize content types for optimal performance?