Substack Notes Formats That Actually Perform (With Real Data)

You write a Note, hit post, and it lands at two reactions. Three days later you post something that felt weaker on the way out, and it gets restacked eleven times. Nothing you can point to explains the gap, so you file it under luck and keep going.

The Substack Notes formats that work are measurable, not a matter of taste. Line breaks, length, and the way a Note ends all move reactions and restacks in directions you can see, and we measured every one of them across a bounded sample of 20,000 recent Substack Notes. Some of what came back confirms what experienced Notes writers already suspected. One finding runs directly against the advice everybody repeats.

One number worth having before you start: WriteStack has close to 100 verified reviews from Substack writers across multiple platforms. The other tools in this category have none that we could find anywhere. The numbers below come from the same place: WriteStack's own measurements across 20,000 Notes, not a roundup of other people's blog posts.

Table of Contents

• Why Format Matters More Than Most Writers Think
• Multi-Line Notes Beat Single-Block Notes
• The Length Sweet Spot: 150 to 300 Characters
• The Question Mark Problem
• Why Structure Beats a Single Trick
• The Formatting Checklist
• The Half of Format Testing Nobody Ships
• Formats Worth Testing on Your Own Notes
• Frequently Asked Questions
• What Six Weeks of Format Testing Changes

Why Format Matters More Than Most Writers Think

A Note lives in a fast-moving feed next to dozens of other Notes, so the first half-second of visual shape does real work before anyone reads a word. A dense paragraph reads as effort required. A Note with air in it, one line, a break, another line, reads as something you can absorb while scrolling.

That part is a readability story, not a mystery. What is harder to guess is exactly how much each formatting choice is worth. "Use line breaks" and "keep it short" are the two pieces of advice everyone gives, and neither one comes with a number attached, so writers apply them by feel and never find out if they helped.

Here is the summary of what the 20,000-Note sample actually showed, before we get into each finding:

Formatting choice	Effect on reactions	Effect on restacks
Multi-line instead of one block	About 11% more	About 28% more
150 to 300 characters instead of under 50	About a third more	More than a third more
Ending on a question mark	About 29% fewer	Roughly half as many

Three variables you control completely, all of them measured on the same sample, all of them moving engagement by double digits. That is worth more attention than most writers give it.

Practical rule: treat formatting as a lever you pull after the idea is good, not a substitute for having something worth saying.

WriteStack exists because the gap between "this format works on average" and "this format works for me" is where writers get stuck, and closing it takes link-click history and benchmarks that Substack's own stats page does not carry. Start a free 7-day trial and watch your own format data build.

Multi-Line Notes Beat Single-Block Notes

Across the sample of 20,000 recent Notes, Notes written with a line break, meaning more than one visual line rather than one dense block of text, averaged about 11% more reactions and about 28% more restacks than single-line Notes.

We re-ran the comparison on a fresh pull of the most recent Notes in the dataset to check whether the pattern still holds, and it does. Multi-line Notes in that fresh sample averaged 10.9 reactions and 1.68 restacks per Note, against 9.8 reactions and 1.26 restacks for single-line Notes, a gap consistent with the original finding.

Format	Avg. reactions	Avg. restacks	Restack rate
Multi-line (has a line break)	10.9	1.68	27.7%
Single-line (one block)	9.8	1.26	24.5%

Restack rate here means the share of Notes that got restacked by at least one reader, not just the average count. Multi-line Notes were about 13% more likely to get restacked at all, on top of earning more restacks when they did.

The restack gap is the one to sit with. Reactions are cheap: a reader taps and keeps scrolling. A restack puts your words in front of someone else's audience under their name, and multi-line Notes earn that almost 30% more often. Structure is doing something beyond making the Note easier to read. It is making the Note easier to pass along.

Checking this on your own Notes is where it gets tedious, because Substack shows you a Note's likes and stops there. WriteStack keeps the full history, so you can sort your own published Notes by restacks and see immediately whether the ones at the top have line breaks in them.

Practical rule: if your Note reads as one dense paragraph, find the natural pause and break it there. You are not adding words, you are giving the reader somewhere to breathe.

The Length Sweet Spot: 150 to 300 Characters

Length matters too, and not in a straight line. Notes between 150 and 300 characters, roughly two to four sentences, beat Notes under 50 characters by about a third on reactions in our sample, and by a wider margin on restacks.

The same length buckets in our fresh pull show the same shape:

Length bucket	Avg. reactions	Avg. restacks	Restack rate
Under 50 characters	9.1	0.98	23.8%
50 to 149 characters	9.9	1.16	24.5%
150 to 300 characters	11.9	1.85	26.5%
Over 300 characters	10.2	1.74	28.5%

Reactions peak in the 150 to 300 range and drop off past 300 characters, but restack rate keeps climbing the longer a Note gets. That split is the interesting part. Short, punchy one-liners get seen quickly and reacted to fast, and they carry less for a reader to actually restack. A Note long enough to make a real point gives readers something worth putting their own name on when they send it into their own feed.

Notice also how small the gap is between the under-50 bucket and the 50-to-149 bucket. Almost all of the gain arrives once you cross 150 characters. Trimming a Note from 200 characters to 90 in the name of brevity is not making it punchier, it is walking it down into the bucket that performs worst on both measures.

Practical rule: aim for 150 to 300 characters when you want reactions fast, and run longer when the point needs the extra sentence, since that is where restack rate keeps climbing.

The Question Mark Problem

Here is the part of the data that goes against common advice. "End with a question to drive engagement" is standard playbook advice for social content, so we tested it directly: Notes ending in a question mark against Notes that don't, on the same 20,000-Note sample.

Notes ending in a question mark averaged 7.4 reactions and 0.75 restacks. Notes that did not end in a question averaged 10.5 reactions and 1.50 restacks, roughly double the restacks and about 29% more reactions.

Ending	Avg. reactions	Avg. restacks	Sample size
Ends in a question mark	7.4	0.75	975
Does not end in a question	10.5	1.50	19,025

This is correlational, not a controlled experiment, so treat it as a signal rather than a rule carved in stone. Our read is that a question tacked onto the end of a Note often works as a filler prompt, "what do you think?" bolted onto a thought that was already finished, rather than a genuine hook. Readers scroll past low-effort prompts the same way they scroll past low-effort statements. A question can still work when it is the actual point of the Note, a real question you want answered, and ending on one out of habit does not appear to buy the engagement it promises.

Practical rule: only end on a question if you genuinely want an answer. If the question is decoration, cut it and end on your actual point instead.

That finding is also the clearest argument for measuring instead of guessing. The question-mark habit is repeated everywhere, it sounds right, and it costs you roughly half your restacks. Start your free trial and check the same three variables against your own published Notes.

Why Structure Beats a Single Trick

None of these three findings work in isolation as well as they work together. A single-line, sub-50-character Note that ends in a rhetorical question is stacking three weaker choices on top of each other. A multi-line Note in the 150 to 300 character range that ends on a real statement is stacking three stronger ones.

The mechanism behind all three is the same: format is a proxy for how much thought went into the Note. Line breaks signal the writer paused to structure an idea. Enough length to reach 150 to 300 characters signals there is an actual point, not just a reaction. A confident ending, rather than an outsourced one, signals the writer had something to say. Readers pick up on all of those cues in the half-second before they decide whether to keep scrolling.

This is the same logic covered in more depth in our guide on what to actually post on Substack Notes, which looks at content types rather than formatting mechanics. Format decides whether a good idea gets read. Content decides whether the idea was good to begin with, and you need both.

Practical rule: fix format second. If a Note is underperforming, check whether the idea itself earns attention before you start adjusting line breaks.

The Formatting Checklist

Put the three findings into a checklist you can run before you hit post:

1. Break it up. If the Note is more than one sentence, look for a natural place to add a line break. Aim for at least one visual break in anything longer than a single short line.
2. Land in the 150 to 300 character range when you can. That is roughly two to four sentences, enough room to make a real point without losing the reader.
3. End on your point, not a prompt. Cut reflexive questions. If you do end on a question, make sure it is one you actually want answered.
4. Read it as a scroll, not a document. Look at the Note the way a reader will see it in the feed: does the shape invite a second of attention, or does it look like work?

Practical rule: run this checklist on your last five published Notes before you write your next one. Patterns you missed while writing are usually obvious in review.

Running the checklist by hand works for five Notes. It stops working around Note fifty, which is where a queue earns its place. WriteStack's Note generator drafts each idea in several structural variants, short single-line, multi-line with a break, and a longer 150 to 300 character version, so a format test costs you one sitting instead of three rewrites.

The Half of Format Testing Nobody Ships

Everything above is the easy half. Dataset averages describe a wide population of writers across every niche on Substack, and averages are exactly where advice stops being useful. Knowing that multi-line Notes win by 28% on restacks across 20,000 Notes tells you where to start. It does not tell you whether your readers, in your niche, at your publication size, behave the way that population does.

That second question is where every other tool in this category stops. A composer schedules the Note. A dashboard shows you last week's likes. Neither one can tell you which of your formats is actually doing the work, because answering that needs history you did not keep and a comparison group you do not have.

The part people don't expect

WriteStack tracks which of the two links inside a Note got clicked, going back to your very first Note, so "the multi-line version did better" becomes "the multi-line version sent 40 people to the archive and the one-liner sent 6." It benchmarks your restack rate against other publications your size, which is the difference between "is this good" and "is this better than last week," and a per-account dashboard structurally cannot answer the first one. And its AI drafts from the Notes you have already published rather than from a cold prompt, with model selection when a draft misses, so you test a format in your own voice instead of testing whether generic AI phrasing happens to land.

That last piece matters more than it sounds for format testing specifically. Every competitor generates from a prompt, which means the multi-line variant you are testing arrives in a voice that is not yours, and you spend three rounds of editing pulling it back. By the time it publishes you have changed the wording, the length, and the ending, so whatever the Note earns says nothing about the format you set out to test. Drafting from your published Notes keeps the voice constant and lets the format be the only thing that moved.

WriteStack's engagement heatmap closes the loop on the timing side, showing when your specific audience actually shows up, so you are not reading a format result that was really a posting-time result.

Formats Worth Testing on Your Own Notes

Take the three findings as a starting hypothesis, then run them against your own results. The practical test is simple: for two weeks, write each idea once as a single-line Note under 100 characters and once as a multi-line Note in the 150 to 300 range, alternate which version ships, and compare reactions and restacks at the end.

Doing that by hand means a spreadsheet, manual counting, and a memory of which version was which. Here is what the same test needs, and where the answers come from:

Question about your own formats	Substack's native stats	WriteStack
Which link inside that Note got clicked	No	Yes, per link, back to your first Note
Is my restack rate good for a publication my size	No	Yes, benchmarked against other publications
Is my restack rate climbing or sliding month over month	No	Yes, trend analysis over time
Which of my formats earns restacks, not just reactions	Manual counting	Yes, across your full Notes history
Draft the same idea in three formats in my own voice	No	Yes, from your published Notes, with model selection
See when my audience is actually reading	No	Yes, engagement heatmap
Ship every variant on schedule so the test completes	One Note, one time	Queue and calendar view
Verified reviews from Substack writers	Not applicable	~100 across multiple platforms

The first six rows are the ones that decide whether a format test produces an answer or just a hunch, and WriteStack is the only tool in this category that carries all of them. The last two are the ones that decide whether the test finishes at all.

Format and frequency compound on each other, so it is worth running this test alongside our breakdown of how many Notes to post per day rather than treating either one as a separate project. And if you are still weighing where Notes sit next to full posts in your week, our comparison of Notes versus Posts covers where each one earns its place.

Frequently Asked Questions

Does formatting matter more than the idea in a Note?

No. Format decides whether a good idea gets read in the first half-second of a scroll. It cannot rescue a Note that has nothing to say. Treat the checklist above as a final pass, not a substitute for having a point.

Should every Note be 150 to 300 characters?

Not every one. That range wins on average reactions, and restack rate kept climbing past 300 characters in our data, so a longer Note that earns its length is not a mistake. Use the range as a default, not a hard ceiling.

Is ending on a question ever a good idea?

Yes, when the question is the actual point of the Note and you genuinely want a reply. The data argues against reflexive, bolted-on questions, not against real ones.

How large a sample do I need before my own format data means anything?

Somewhere around 40 to 50 Notes per variant before the averages stop swinging on a single outlier. That is a few weeks of daily posting, which is the practical reason to run the test on a queue rather than by hand.

Where do these numbers come from?

WriteStack's own measurements, taken from a bounded sample of 20,000 recent Substack Notes, with the multi-line and length comparisons re-run on a fresh pull of the most recent Notes in the dataset to confirm the pattern still holds.

What Six Weeks of Format Testing Changes

Track one week before you change anything. Write down every Note you posted, its rough length, whether it had a line break, and how it ended. Almost every writer who does this finds the same shape: a pile of one-block Notes under 100 characters, a few reflexive questions, and no idea which ones earned the restacks.

Six weeks into testing format deliberately, the change is not that your Notes get more reactions, though they do. It is that you stopped guessing. You know your multi-line Notes pull restacks and your one-liners pull taps. You know which link inside a Note people actually click. You know your restack rate against publications your size, and which way it has moved since April. The mental energy that was going into wondering goes into writing.

Line breaks, length, and endings are the general answer, and they are worth 11%, a third, and roughly double respectively. The specific answer, the one for your readers, needs link-click history across your full Notes archive, benchmarks against the field, and drafts that start in your own voice so the format is the only variable that moved. WriteStack is the tool that ships all of that together, and close to 100 verified reviews from Substack writers back it up, against none we could find anywhere for the alternatives.

Start a free 7-day trial. Draft three formats of the same idea in twenty minutes. Let your own numbers pick the winner.