Thread structures that keep people reading: six patterns

A good thread has a shape the reader can feel: they know why they're reading, what comes next, and when it ends. Pick one of six patterns before you write, and most of the hard decisions are made for you: what goes in part 1, how to order the rest, and how to finish.

This post is about the shape of a whole thread. For the opening line, see writing hooks that stop the scroll. For splitting a text into parts, numbering and length, use the thread splitter. Limits per post are on the limits cheat sheet.

How to choose a pattern

Ask what the reader should have at the end:

You want the reader to…	Use
Follow a sequence of events	Story
Pick up several ideas fast	List
Change a belief	Myth vs fact
See a difference	Before and after
Trust your method	Case study
Do something themselves	How-to

You can mix patterns, but one thread should have one main promise.

What every part does

Whatever the pattern, each part has a job:

• Part 1 promises. It says what the thread gives the reader, so they can decide to open it.
• Middle parts deliver. One idea each, and each should still make sense if it's the only part someone sees in their feed.
• The last part closes. A summary, a question, or one next step. Don't end on "thanks for reading".

1. Story

Parts follow time: a situation, a problem, what you tried, what happened, what you learned.

• Part 1: start at the moment of tension, not at the background.
• Middle: one turn of events per part. Cut the details that don't change what happens next.
• Last part: the lesson, in one sentence, and a question for people who've been there.

Weak spot: background that runs for two parts or more before anything happens. Move the problem forward.

2. List

A promise of N items, then one item per part.

• Part 1: name the list and who it's for: "N things I'd tell someone starting X".
• Middle: each item is a claim plus one line of why or how. Put your strongest item first or last, not in the middle.
• Last part: the item readers will argue with, or a request: "What's missing?"

Weak spot: filler items. If an item doesn't earn its place, remove it and change the number.

3. Myth vs fact

Each part takes one common belief and corrects it.

• Part 1: say which belief you're taking apart and that you'll back each correction.
• Middle: "Myth: …" then "What actually happens: …" with the reason. Link a source for any fact about Threads, preferably Meta's own pages.
• Last part: the one thing to do differently.

Weak spot: a "myth" that nobody actually believes. Pick beliefs your audience really holds.

4. Before and after

Show the same thing twice, then explain the difference.

• Part 1: the "before", shown as it was.
• Part 2: the "after".
• Middle: the two or three changes that mattered, one per part.
• Last part: how the reader can apply the same changes to theirs.

Weak spot: an "after" that is only claimed. Show it, using a real draft of your own.

Illustrative example (written for this article, not a real post):

Before: I've been thinking a lot about how I plan my week lately, and I wanted to share some thoughts on what's been working and what hasn't.

After: Three planning habits that make my week easier to manage, and the one I dropped.

The changes: the "after" opens with the promise instead of a warm-up, says how many parts are coming, and hints at a twist (the habit that was dropped) that gives a reason to read to the end.

5. Case study

Something you did, with the evidence that you did it.

• Part 1: what you did and what you measured, in one sentence.
• Middle: the setup, what you changed, what happened.
• Last part: what you'd do differently and what it doesn't prove.

Only use numbers you can show, with the date you measured them. If you have no real result yet, write a story instead.

Illustrative example (a template, not a real result; the brackets are for your own numbers):

Part 1: For four weeks I posted at the same two times every day and noted the replies. Here's what changed, and what it doesn't prove.

Part 2: The setup: same topics, same length, only the times changed. I measured replies after 48 hours, from my account's insights on [date].

Part 3: What happened: [your numbers, before and after, with the dates you measured them].

Part 4: What it doesn't prove: four weeks is short, and topics vary week to week. Next I'll test one time slot at a time.

6. How-to

Steps in the order the reader does them.

• Part 1: the outcome and how long it takes. Say only what's true for most people.
• Middle: one step per part, starting with a verb. Name the one mistake people make at that step.
• Last part: how to check it worked.

Weak spot: a step that needs something the reader doesn't have. List what they need in part 1.

Before you publish

1. Read only part 1. Would you open the thread?
2. Read each middle part alone. Does it still make sense?
3. Remove the part you like least. Is the thread better?
4. Check the last part ends with one clear thing.

Then split the text with the thread splitter and check each part's length with the character counter. In Govir, write the parts in Write with Add to thread, then Schedule: Govir publishes every part in order at the time you pick (Scheduling & Calendar).

After a few threads, compare which pattern brought the most replies and follows, and write more of that. Change one thing at a time so you know what made the difference.