Too long? Too short? Analyzing my Substack post lengths
🗣️ How looking at variability in size and effort for my Substack posts is changing how I plan and publish what I write (audio now available; 5:56)
This post was originally published on agileteams.substack.com on May 6, 2024. As of July 12, all agile Teams content has moved to the agile section of karensmiley.substack.com. Please join us there!
Planning by Post Sizes
Thanks in part to a post by
about his thoughts on post sizes, I realized in my March retrospective that I need to plan writing effort on my posts based on sizes. But what’s “short” and what’s “long” for me and for my readers?To evaluate size thresholds, I looked through two lenses: my own data, and industry practices. This post summarizes what I found, and what I’m going to do next 🙂
Thanks for reading agile Teams! Subscribe for free to receive new posts (about monthly) and support my work.
What my data says
Analyzing my post lengths for March and April shows distinct groupings, with a big gap between shorter and longer posts (2000-3000 words):
Using smaller buckets showed an additional gap between 600-900 words:
Seeing this, I drilled into the big clump of posts under 600 words to check for clusters:
I didn’t drill further because this is obviously a small sample (29 posts in total). I did notice a small increase in post lengths from March to April, but it’s way too soon to identify any trends or shifts.
(Yes, I could have analyzed the groupings with ML, but this isn’t big data. It’s very small data, I just did the “simplest thing that could possibly work”, and groupings were visually obvious.)
What the internet says
Out of curiosity, I did some quick searches for wisdom on recommended post lengths. One focused on blogging suggests 275 words for maximum engagement when sharing links, quotes as stale advice about keeping posts under 600, suggests 600-1500 for maximum shares on social media, and recommends “longer, heavily researched posts 2,450+ words long”1 for maximum reach on Google. Their recommended categories:
“Micro content: 75-300 words
Short Form content: 300-600 words
News Article length content: 750 words
Mid-form content: 1000-1500 words
Long-form Content: 2450+ words”
(The gaps in categories between 600-1000 (except for 750) and between 1500-2450 are interesting 🤔.)
Another article, by Hootsuite2 , recommends:
25 words (!) for LinkedIn posts, and
1900-2000 words for LinkedIn articles.
(They cover Facebook, Twitter, Instagram, TikTok, LinkedIn, Youtube, Pinterest, and Snapchat, but not Substack or Medium.)
Decision
Since Substack writing isn’t blogging, I’m not going to stick strictly with those guidelines. Also, I’m currently counting the words in my end notes in the totals, which might not be consistent with how they are counting (can’t tell). The 2450+ “heavily researched” post description fits pretty well with my Long articles and their many end notes, though - that feels validating. 👍
Bottom line: I’ve decided to group my posts in 4 size categories to help me with planning and estimating my writing effort:
Very Short (< 200 words)
Short (200-799 words)
Medium (800-2499 words)
Long (>2500 words)
These annotated graphs show where the 4 categories fit against my data to date:
As is visible on the graph, my longest post was over 5000 words. I considered splitting the Long category (2500+ words) to add an XL category (5000+ words). However, I’m feeling like the super-long posts might take disproportionately longer to finalize, and I might be better off to break them up into multiple posts (if I can find a division that makes sense logically for the topic).
Next Steps
For now, I’m going to keep one open-ended Long category. During May, whenever a draft post gets above 3000-4000 words, I’ll focus on making a deliberate decision on whether the post should be split. And I’ll re-evaluate the groupings in early June, as part of my next retrospective, after my May posts are completed.
This approach is going to be tested immediately. My
Part 3 draft on ethics of genAI for music was already 6833 words as of April 26, the day when I carved off the “AI Ventriloquism” bonus topic to a separate article (which ended up being just over 3000 words itself). And I’m returning to that Part 3 draft post as soon as my month 2 retrospective is published … 🙂I’d love to hear from anyone else who is looking at data on their Substack post sizes, and what your experiences are with effort and engagement on shorter and longer posts!