Sort

SORT

SORT is what you use when your material is a mess and you want ChatGPT to stop guessing and start handling it like a competent assistant. Not magic, not mind reading, just a process.

This template is especially useful when you are working with notes, transcripts, pasted documents, meeting summaries, brainstorm lists, or any other pile of text that looks like a desk drawer from 1997.

What the template means

Letter or word	Prompt ingredient	What it does
S, Separate	Split the material into parts	Helps ChatGPT identify categories, themes, sections, or types of information instead of treating everything like one blob
O, Organize	Put the parts into a usable structure	Turns scattered content into groups, headings, tables, steps, or outlines
R, Rank	Prioritize what matters most	Tells ChatGPT to compare items and sort them by importance, urgency, quality, or relevance
T, Transform	Convert the material into a new format	Changes raw input into something useful, like action items, summaries, study notes, FAQs, or an email draft

Why this template works

SORT helps because most bad prompts fail for a simple reason: they throw in material and ask for a result without giving a process.

If you paste a transcript and say, “summarize this,” ChatGPT will do something. Maybe even something decent. But if you ask it to separate, organize, rank, and transform, you are giving it a workflow.

That matters.

SORT is part of working with material, which means you are no longer asking random one-off questions. You are asking ChatGPT to handle actual content. Notes from a meeting. Customer feedback. Research quotes. Brainstorm ideas. Training documents. The messy stuff people actually have.

Here is what SORT adds:

Separate reduces confusion.
Organize creates structure.
Rank adds judgment and priority.
Transform produces a useful final output.

So instead of getting a vague summary, you get something you can actually use, share, or act on.

3 sample prompts

Sample prompt 1: Formal version

SEPARATE: Divide the meeting notes below into these categories: decisions, action items, open questions, risks, and follow-up topics.
ORGANIZE: Present each category as a heading with bullet points underneath.
RANK: Within action items, rank tasks from highest to lowest priority based on urgency and business impact.
TRANSFORM: Turn the final result into a professional project update summary for a department manager.

Meeting notes:
We agreed to delay the website launch by one week because the payment system still has testing issues. Maria will follow up with the vendor by Thursday. Customer support needs a new FAQ before launch. There is still disagreement about the refund policy wording. Legal wants one more review. The design team finished the mobile homepage, but desktop banners are still pending. We may need temporary copy if final product descriptions are late.

How the template shows up:

S, Separate: The prompt tells ChatGPT exactly which categories to split the notes into.
O, Organize: It specifies headings with bullet points, so the output has structure.
R, Rank: It asks for action items to be prioritized by urgency and business impact.
T, Transform: It converts messy notes into a professional update summary.

Why it improves the prompt: This version removes ambiguity. ChatGPT does not have to guess what counts as important or how to format the answer. It follows a clear sequence, which usually leads to cleaner, more usable output. This is the prompt equivalent of handing someone labeled folders instead of dumping papers on their chair.

Sample prompt 2: Semi-structured version

I pasted customer feedback from survey responses below.

First, separate the comments into themes such as product quality, shipping, customer service, and price.

Then organize the themes into a short report with clear headings.

Rank the themes by how frequently they appear and how serious the complaints seem.

Finally, transform the report into 5 recommendations for the operations team.

Feedback:
“Shipping took forever.”
“The product works great, but setup instructions were confusing.”
“Customer service fixed my issue fast.”
“The price feels too high for what you get.”
“Delivery box arrived damaged.”
“Support was polite, but I had to wait too long.”
“I like the product quality.”
“Instructions need pictures.”

How the template shows up:

S, Separate: The comments are split into themes like shipping, service, and price.
O, Organize: The themes are arranged into a short report with headings.
R, Rank: The themes are prioritized by frequency and seriousness.
T, Transform: The findings are converted into recommendations for a team.

Why it improves the prompt: This version is less rigid, but still gives ChatGPT a clear path. It is especially useful when you want insight, not just summarization. The ranking step is the secret sauce here, because it helps move from “here is what people said” to “here is what deserves attention first.”

Sample prompt 3: Natural version

I pasted a messy list of ideas for a workshop. Please split them into related groups, put them into a logical order, tell me which ideas seem most useful for beginners, and then turn the whole thing into a simple 45-minute workshop outline with a title, 3 sections, and a short closing activity.

How the template shows up:

S, Separate: “Split them into related groups” handles the separation step.
O, Organize: “Put them into a logical order” gives structure.
R, Rank: “Tell me which ideas seem most useful for beginners” asks for prioritization.
T, Transform: “Turn the whole thing into a simple 45-minute workshop outline” changes the material into a final format.

Why it improves the prompt: This one sounds natural, which is good news if you do not enjoy writing prompts like legal documents. Even without explicit labels, the SORT process is still there. The result is likely to be more practical than just saying, “make this into a workshop.” You are guiding the model through the messy middle, where most useful work actually happens.

Warnings and extra tips

SORT is strong, but it is not psychic.

A few things to watch out for:

Give the material. SORT works best when you actually provide notes, text, comments, or ideas. Without input, ChatGPT will fill gaps with assumptions.
Define the ranking rule. “Rank these” is decent. “Rank these by urgency, cost, and customer impact” is much better.
Do not skip the transform step if you need an end product. Many people stop at organization and then wonder why the output still feels unfinished.
Be realistic about messy source material. If your notes are incomplete, contradictory, or chaotic, the answer may still need human review. Sorry, the template cannot perform archaeological miracles.
Avoid overloading one prompt. If the source material is very long, break the task into chunks first. Otherwise you may get shallow sorting instead of thoughtful analysis.
Check the categories. If you tell ChatGPT to separate content into bad categories, it will obediently produce bad organization. Very efficient, very wrong.

A good upgrade is to ask ChatGPT to show uncertainty. For example: “If any item fits more than one category, note that.” That can make the output more honest and more useful.

Quick summary

SORT gives ChatGPT a practical workflow for handling messy material:

Separate the pieces
Organize the structure
Rank the priorities
Transform the result into something useful

That is why SORT works so well for transcripts, notes, documents, survey feedback, and brainstorms. It turns chaos into output you can actually use.

Try it next with something real from your own life, a meeting summary, a page of notes, customer comments, or a rough outline. Paste it in, run SORT, and see how much better the result gets when you stop asking for “help” and start asking for a process.