Back to AI Basics – Top Prompt Strategies & Techniques

AI tools are getting smarter, but the quality of the result still depends heavily on the information and direction you provide. In this Member Training, Chris breaks down the prompting habits that produce better, more accurate, and more useful results in ChatGPT and Claude.

You will see live comparisons between vague and detailed prompts, learn how context changes the final product, and discover why examples are one of the most effective ways to guide AI. Chris also demonstrates how these techniques can be used to create social posts, market reports, listing descriptions, newsletters, property flyers, and other real estate marketing materials.

What You’ll Learn

• How to write clear, direct prompts
• Why context can be more important than lengthy instructions
• How to give AI the right audience, purpose, goals, and background
• How one-shot and few-shot prompting improve consistency
• How to use written and visual examples
• How to specify the format and structure of the final result
• How to refine an answer through follow-up prompts
• How to ask AI for sources and verify its claims
• When to use different AI models
• How projects, memory, files, connectors, and skills add context
• Why AI should assist your expertise, not replace your judgment

Important Timestamps

00:00 | Welcome and Session Overview
Chris explains how prompting has changed as ChatGPT and Claude have become more capable. He also introduces the written session notes members can use while practicing these strategies.

07:30 | Why Good Prompting Still Matters
See how AI can produce a polished and convincing answer that is still inaccurate. Chris revisits the example of a seller using ChatGPT to estimate a home’s value and shows how the answer changes once the AI is questioned.

13:15 | Strategy 1: Be Clear and Direct
Learn why vague instructions often produce incomplete or generic results.

15:00 | Live Open House Social Post Comparison
Chris compares a basic “write a social media post” request with a stronger prompt that includes tone, purpose, selling points, property details, and the CMA or MLS information.

22:00 | Let AI Ask Clarifying Questions
Newer models are becoming better collaborative partners. See why allowing AI to stop and request missing information can prevent it from inventing details.

26:15 | Strategy 2: Give It Context
Chris explains how telling AI who you are, who the content is for, why you are creating it, and what you want to accomplish can dramatically improve the result.

29:00 | Market Report With and Without Context
Compare two versions of a Circle C Ranch market report. One reads more like a formal webpage, while the contextual version is better organized for a newsletter sent to an agent’s sphere.

37:30 | Context Engineering and Voice Dictation
Learn how files, photos, spreadsheets, projects, memory, plugins, connectors, and previous conversations contribute to the AI’s context. Chris also shows how voice dictation can be used to provide detailed background quickly.

40:00 | Building a Personal CRM With Context
Chris shares an example of how he used Claude to build a personal CRM around his actual goals, Gmail, iMessages, contacts, and follow-up needs.

44:45 | Strategy 3: One-Shot and Few-Shot Prompting
Learn how giving AI one or more examples helps it recognize and reproduce the tone, structure, style, and pattern you want.

47:00 | Listing Description Demonstration
Chris uploads property information and provides several sample listing descriptions. Claude uses those examples to create a new description that follows the established style.

51:00 | Choosing the Right AI Model
More power is not always better. Chris explains why simpler tasks may work better with a faster model and how using the most advanced model for everything can waste tokens or cause the AI to overthink the assignment.

54:45 | Creating Marketing From a Visual Example
Chris takes a screenshot of a Canva property flyer, uploads listing photos and property information, and asks ChatGPT to recreate the design for a different property.

55:00 | Additional Prompting Techniques
This section covers:

• Specifying the required output format
• Iterative and conversational prompting
• Prompt chaining
• Using AI to draft a reusable system prompt
• Claude’s Meta Prompt
• A practical “when in doubt” prompt framework

59:00 | Member Questions and Troubleshooting
Chris begins the live question-and-answer portion and explains how to improve or challenge AI-generated responses.

1:03:30 | Reviewing the AI-Generated Property Flyer
See the completed flyer based on the Canva example. Chris reviews what it handled well, what still needs refinement, and how follow-up prompts can improve it.

1:07:15 | Requesting Sources and Citations
Learn how to ask AI to cite its sources within a market report. Chris also explains why you should review the sources and instruct the AI not to rely on a competitor’s website.

1:10:45 | Role Prompting
Chris discusses prompts such as “act as a real estate professional,” when a role can help, and why assigning a role does not automatically give AI genuine professional expertise.

1:13:30 | Adding Images and Graphics to Newsletters
Learn how a newsletter design or template can be used as a visual example, along with your preferred images and written content.

1:15:00 | Editing AI-Generated PDFs in Canva
Chris demonstrates how to download a PDF created by AI, upload it to Canva, and convert the text and images into editable elements.

1:17:00 | Connecting AI to Brevo and Other Platforms
This section explains connectors, plugins, MCP servers, API keys, and some of the limitations and risks involved when allowing AI to interact with outside platforms.

1:22:00 | Prompting Resources and Practical Limitations
Chris discusses the updated session notes, why an earlier AI-generated cheat sheet was not accurate enough to share, and why human expertise remains necessary.

1:24:00 | Final Takeaways
The session closes with a reminder to give AI the information it needs, ask better questions, verify its responses, provide examples, and continue refining the result.

A Simple Prompt Framework

When you are unsure how to begin, include these six elements:

1. Who you are: “I am a real estate agent working in…”
2. The task: Clearly state what you want AI to create.
3. The audience: Explain who will read or use it.
4. The purpose: Describe what you want the content to accomplish.
5. The information: Provide the property details, market data, files, examples, or background.
6. The format: Specify the desired length, tone, sections, structure, and final file type.

Key Takeaway

AI can create an impressive first draft, but impressive does not always mean accurate. Give it clear instructions, meaningful context, and strong examples. Then question, verify, and refine what it produces using your own real estate knowledge and professional judgment.