Common Substack Notes Mistakes (Backed by Real Data)

Published: September 2026

Most "mistakes to avoid" lists are opinions. This one leans on an actual dataset: an analysis of 9,641 Notes from 689 writers, posted in April 2026, measured against the metric that actually matters — free subscribers gained per Note, not likes or clicks (Write Build Scale). A few of these findings will contradict generic advice you've probably read elsewhere, including some of ours — worth knowing which advice is actually measured versus assumed.

1. Skipping a credentialing hook

Notes that opened with a specific, verifiable credential averaged 3.13 subscribers each, versus 0.15 for Notes without one — roughly a 20x difference. Even after removing a few heavy outliers, the effect held at a 2.1x lift. A vague opener costs you the single highest-leverage thing you can put in the first line.

2. Not using repetition (anaphora) in the opening lines

Notes that repeated the same opening word or phrase across consecutive lines — 408 of the Notes studied — averaged 0.48 subscribers versus a 0.15 baseline, and hit a 13.7% viral rate versus 9.7% for Notes without it. A structural device, not a content idea, moved the number meaningfully.

3. Writing too long

The sweet spot in this dataset was 31–60 words, averaging 0.22 subscribers and an 11.3% viral rate. Notes in the 61–120 word range dropped to 0.13 subscribers — consistent with other research on Notes length, but with an exact number attached this time.

4. Ending with a question — at least for subscriber conversion specifically

This is the one worth pausing on: Notes containing a question mark averaged 0.116 subscribers versus 0.179 for question-free Notes — a 35% reduction, and the penalty was steeper (59%) for personal-story-style Notes specifically. This doesn't mean questions are bad — they're still a solid way to drive replies and comments, which matters for algorithmic reach. It means a question mark measurably underperforms on the specific metric of converting a reader into a subscriber. Those are two different goals, and this dataset only measured one of them.

5. Chasing likes over restacks

Restacks converted at 0.0429 subscribers each, versus 0.0035 for likes — a 12x difference in conversion power. A Note optimized to be liked and a Note optimized to be restacked aren't the same Note; if growth is the goal, restack-worthy (something someone wants to put in front of their own audience) beats merely agreeable.

6. Ignoring which day you post

Sunday Notes in this dataset averaged 0.236 subscribers versus 0.133 for Wednesday — a 78% gap. This is one dataset, not a universal law (see our own timing research for why audience-specific testing still matters more than copying someone else's best day) — but it's a real enough gap to be worth testing on your own account rather than assuming every day performs the same.

7. Giving up before the algorithm has enough data

A separate account of writers stuck around 100 subscribers (Escape the Cubicle) points to abandoning Notes after 2–3 weeks as a common, specific mistake — the recommendation there is a 90-day minimum at 2–3 Notes a day, since the algorithm needs a real sample of your Notes to learn who to show them to, not a handful.

8. Writing dense blocks of text

The same source found identical content reformatted with short sentences and line breaks got roughly 3x the engagement of the same words as a dense paragraph. On a feed people scroll quickly, formatting is doing real work, independent of what the Note actually says.

9. Publishing and disappearing

A different category of mistake entirely, since it's about behavior around the Note rather than the Note itself: posting and then vanishing until the next one, with no replies to comments and no presence in the hours after. This reads to readers as a transaction, not a relationship — and Substack's own engagement research treats it as one of the biggest, most fixable mistakes new writers make (Mack Collier). The suggested fix is concrete: block 30 minutes right after publishing specifically for replying to comments and engaging with anyone who restacked.

10. Writing Notes but never commenting on anyone else's

Comments on other people's Notes are reportedly more visible than the Notes themselves, which makes skipping this step a real missed opportunity, not just an etiquette lapse. One specific ratio worth trying: three genuine comments on other people's Notes for every one Note you write yourself. This connects directly to the audience-overlap mechanism in Substack's own algorithm — showing up in other people's comment sections is part of how that overlap gets built in the first place.

11. Spending engagement energy on the wrong accounts

A specific misallocation: spending most of your engagement time replying to and restacking large, well-known accounts, whose audiences are too big to notice one more comment from a new writer. The better return is reportedly the opposite — roughly 80% of engagement time spent with writers at a similar subscriber level, who are actually likely to notice, reciprocate, and recommend you back.

The pattern across all of this

The first eight mistakes above are measurable inside a single Note — a credential, a word count, a punctuation mark, a day of the week. The last three are about what happens around your Notes, not inside them — whether you show up after publishing, whether you engage with peers before expecting them to engage with you. Both categories matter, and neither is fixed by getting better at writing alone. That's the actual argument for tracking your own Notes against real data instead of a general best-practices list: your own account's numbers will tell you which of the content mistakes apply to your specific audience, while the relationship mistakes are worth auditing honestly regardless of what your stats say.

Test which of these actually hold for your audience. NotesIQ's Content Type Breakdown and engagement tracking are built from your own real Notes, not someone else's dataset.