Using ChatGPT for Site Plan Design: Practical Prompts That Save Real Time

Let’s be clear upfront: ChatGPT cannot draw a site plan. It’s a text-based AI. It doesn’t produce drawings, CAD files, or spatial layouts. If you’re expecting to type “design me a site plan” and get a buildable result, that’s not how this works.

What ChatGPT is genuinely useful for in site planning is the thinking, research, and documentation work that surrounds the design - the analysis, code research, programme development, design rationale writing, and checklist creation that takes hours and isn’t the creative part of the job.

Here’s how to actually use it.

---

What ChatGPT Can Do for Site Planning

Task	How Useful	Time Saved
Zoning code research and summarisation	Very useful (with verification)	1-3 hours per project
Site analysis checklist generation	Very useful	30-60 minutes
Programme brief development	Useful	30-60 minutes
Design rationale writing	Very useful	1-2 hours
Parking calculation and requirements research	Useful (verify locally)	30 minutes
Sustainability checklist (LEED, BREEAM criteria)	Useful	1 hour
Client presentation text	Very useful	1-2 hours
Drawing a site plan	Not possible	N/A

---

Practical Prompts You Can Use Today

1. Site Analysis Framework

Prompt:

I'm designing a [building type] on a [size] site in [location].
The site is [describe: flat/sloped, urban/suburban, existing structures].
Neighbouring uses include [describe surroundings].

Create a comprehensive site analysis checklist covering:
- Environmental factors (sun path, prevailing winds, drainage)
- Access and circulation (vehicular, pedestrian, service)
- Views and visual impact
- Regulatory considerations
- Site constraints and opportunities

Format as a checklist I can print and take to a site visit.

Why this works: It generates a structured analysis framework tailored to your specific project. You’ll still need to fill in the actual data from site visits and surveys, but having the framework saves 30-60 minutes of setup time.

2. Zoning Requirements Summary

Prompt:

Summarise the typical zoning requirements for a [building type]
in [city/region]. I need to understand:
- Setback requirements (front, side, rear)
- Maximum building height and storey limits
- Floor Area Ratio (FAR) / plot ratio limits
- Parking requirements (number of spaces per unit/sqm)
- Open space / green cover requirements
- Any special provisions for [specific requirement]

Note: I will verify all figures against the actual local code.
This is for initial feasibility only.

Why this works: Zoning research is time-consuming because codes are long, complex documents. ChatGPT gives you a starting framework of what to look for and approximate ranges. Always verify against the actual local planning code - ChatGPT may produce outdated or incorrect specific numbers.

3. Programme Development

Prompt:

I'm developing a programme brief for a [building type] of
approximately [total area] sqm in [location].
The client requires [list key requirements].

Generate a detailed space programme including:
- List of spaces with approximate areas (sqm)
- Adjacency requirements (which spaces should be near each other)
- Access requirements (public, private, service)
- Any outdoor space requirements
- Typical occupancy numbers per space

Present as a table with columns: Space, Area (sqm), Access Type,
Adjacency Notes, Occupancy.

Why this works: Programme development is iterative and benefits from having a comprehensive starting point. ChatGPT generates a reasonable first draft that you then refine based on client discussions and local requirements.

4. Design Rationale and Planning Statement

Prompt:

Write a design rationale for a site plan proposal with these
characteristics:
- Site: [describe site]
- Building: [describe building type, massing, position on site]
- Access: [describe vehicular and pedestrian access strategy]
- Landscape: [describe landscape approach]
- Key design decisions: [list 3-4 key decisions you made and why]

Write for a planning submission audience. Tone should be professional
and evidence-based. Approximately 500 words.

Why this works: Writing design rationales is one of the most time-consuming documentation tasks. ChatGPT produces a solid first draft that you edit for accuracy and project-specific nuance. This typically saves 1-2 hours of writing time.

5. Parking and Circulation Analysis

Prompt:

For a [building type] with [number of units/sqm of floor area]
in [location], calculate:
- Typical parking requirement (based on standard codes)
- Approximate parking area needed (at 25-30 sqm per space including circulation)
- Bicycle parking requirements
- Service vehicle access requirements
- Suggested circulation pattern (one-way vs two-way)

Note any assumptions you're making. I will verify against local code.

6. Sustainability and Compliance Checklist

Prompt:

Create a site design sustainability checklist for a [building type]
project targeting [LEED/BREEAM/local standard].
Cover site-related credits only:
- Site selection and land use
- Stormwater management
- Heat island reduction
- Light pollution reduction
- Transportation and access
- Biodiversity and habitat

For each item, list the requirement and a brief description of how
it's typically achieved on site.

---

Advanced Workflows: Combining ChatGPT With Design Tools

ChatGPT + GIS Data

Run your site analysis in GIS (QGIS, Google Earth Pro, or local planning portals), then paste the findings into ChatGPT and ask it to summarise and identify design implications. This turns raw data into actionable design inputs faster than writing the analysis manually.

ChatGPT + Regulatory Documents

Copy relevant sections of local planning codes into ChatGPT and ask it to:

• Summarise the key requirements that apply to your site
• Identify any conflicting requirements
• Flag items that need further clarification from the planning authority

This is faster than reading entire code documents and helps you identify the sections that matter for your specific project.

ChatGPT + Client Meeting Notes

After a client meeting, paste your rough notes and ask ChatGPT to:

• Organise them into a structured brief
• Identify any gaps or contradictions in the client’s requirements
• Generate follow-up questions for the next meeting

---

What ChatGPT Cannot Do (Be Honest About This)

Task	Why Not
Draw or generate a site plan	It’s a text model, not a spatial/graphic tool
Guarantee code compliance	It may hallucinate specific numbers or cite outdated codes
Replace professional judgement	It can’t assess site-specific conditions it hasn’t seen
Produce accurate calculations	It’s approximate at best - verify all numbers independently
Understand spatial relationships	It can describe adjacencies in text but can’t evaluate if they work spatially
Access real-time local data	It doesn’t know current zoning amendments, recent planning decisions, or site-specific conditions

The golden rule: Use ChatGPT for the 30% of site planning that is text, research, and documentation. Do the 70% that is spatial design, site judgement, and professional expertise yourself.

---

Common Mistakes When Using ChatGPT for Architecture

1. Trusting specific numbers without verification ChatGPT might say “the setback requirement is 6 metres” - this could be right, outdated, or completely made up. Always verify against the actual code.

2. Using AI output as final documentation AI-generated text is a first draft. It needs professional review, project-specific editing, and fact-checking before it goes into any submission.

3. Over-relying on AI for creative decisions ChatGPT can suggest programme arrangements and design principles, but the spatial creativity, site sensitivity, and aesthetic judgement that make good architecture are yours, not the AI’s.

4. Not providing enough context Vague prompts produce generic results. The more specific you are about site, programme, location, and constraints, the more useful the output.

---

Ready to build your design and technology skills? The Archgyan Academy offers courses for architects looking to integrate new tools into their practice.