Substack Notes Formats That Actually Work
Published: September 2026
Ask "should I use images in my Notes?" and you'll find two studies that seem to disagree. One says images are a clear engagement winner. Another says text-only Notes are the most reliably strong format. Both are looking at real data, and both are right - they're just measuring different things. Here's what each actually found, and what to take from the combination.
The case for images: raw engagement
A large-scale analysis of 12.4 million Substack Notes found that Notes with images get 63% more engagement than text-only Notes. A related figure from the same body of research: out of 5.46 million Notes in that sample, only 25% included an image, and that 25% consistently outperformed the other 75%. The gap isn't subtle, and the opportunity is real - most writers on Substack still aren't using images in Notes, which means the ones who do are standing out in a feed where the majority of what a reader scrolls past is plain text.
The case for text-only: consistency
A separate, smaller but more granular study looked across 39 publications and 1,041 Notes, breaking them into 553 text-only Notes, 293 text-plus-image Notes, and 80 text-plus-link Notes. Here, text-only Notes had the strongest within-publication top-quartile rate - meaning that, publication by publication, a well-written text-only Note was the most consistent way to land in that publication's own top-performing tier. Image Notes showed higher raw engagement averages in some slices but were described as more "outlier-sensitive," meaning their strong average was often being pulled up by a handful of viral hits rather than reflecting a reliable, repeatable lift.
How both findings are true at once
The two studies aren't contradicting each other - they're measuring different things. The 12.4-million-Note analysis is comparing image Notes to text Notes on average engagement, and on that measure, images win clearly. The 39-publication study is asking a different question: if you personally write a Note today, which format gives you the most reliable shot at it being one of your own best-performing Notes - and there, well-executed text-only wins because it doesn't depend on having a strong image to pair with a mediocre thought. Put together, the honest takeaway is: images raise your ceiling, but only if the image and the text around it are actually good; text-only is the safer, more consistent floor if you don't have a genuinely useful image to add.
Structure matters more than most people think
Two separate structural findings are worth building into how you actually write a Note, regardless of whether it has an image:
- Line breaks measurably help. Notes written with at least one line break averaged about 11% more reactions and 28% more restacks than single-line Notes. A wall of unbroken text is harder to skim in a feed people are scrolling through quickly, and the data reflects that directly.
- Multi-paragraph structure outperforms a single compressed block. In the 39-publication study, Notes broken into 4-6 paragraphs, and even 7+ paragraphs, outranked compressed single-block formats. This doesn't contradict the practical ~250-word ceiling covered elsewhere on this blog - it's not an argument for writing longer Notes, it's an argument for structuring the words you do write into short paragraphs and line breaks instead of one dense block. A 180-word Note split into four short paragraphs reads faster and performs better than the same 180 words as one unbroken chunk.
Links are the one format that consistently underperforms
Across the 39-publication study, text-plus-link Notes were the clear underperformer, with the analysis concluding plainly that "Notes with a link to a post don't perform well." This matches the much larger 655,056-Note dataset covered separately, which found a 44% engagement drop on link-carrying Notes. Two independent studies landing on the same conclusion from different data is a stronger signal than either one alone: if you're choosing a format, a link is the one to reach for least often.
What this means for the four practical format choices
Putting the findings together into an actual decision framework for a given Note:
- Have a genuinely relevant image? Use it - the ceiling on image Notes is real, and being in the minority of writers using images at all is itself an advantage in a mostly-text feed.
- Don't have an image, or the image would just be filler? Write it as a well-structured text-only Note rather than forcing an image in. A strong text-only Note with line breaks and short paragraphs is the format most likely to reliably perform, per the top-quartile data.
- Tempted to add a link? Ask whether the click is actually the point - if not, leave it out and let the Note stand on its own.
- Regardless of format, break it up. Short paragraphs and at least one line break are close to a free win, adding double-digit percentage lifts in both studies for something that costs nothing but a few extra returns while typing.
The bottom line
There's no single "best" Note format - there's a ceiling format (images, used well) and a floor format (structured text-only), and the data supports using both depending on whether you have something genuinely worth showing. What's consistent across every dataset here is that structure - line breaks, short paragraphs, and avoiding unnecessary links - measurably helps regardless of which format you're working in, and it's the cheapest change to make starting with your very next Note.
NotesIQ breaks down your own performance by format, so you can see whether images, text, or links are actually working for your specific audience.
Related articles
Substack Notes Character Limit: What Actually Works
Substack won't say what the technical limit is, but real analysis of what actually gets read points to a much shorter practical ceiling.
How to Write Viral Substack Notes
What real analyses of viral and failed Substack Notes actually show about hooks, formatting, and the single most common mistake.
Do Links in Substack Notes Help or Hurt Reach?
An analysis of 655,056 Notes found links cut engagement by 44%. Here's exactly when that tradeoff is still worth it, and how to place a link to minimize the damage.