myReading Time Estimator
Paste any text to estimate reading time at slow, average and fast reading speeds.
About the Reading Time Estimator
Displaying estimated reading time on articles and blog posts increases engagement and reduces bounce rate — readers who know a post takes 4 minutes are more likely to commit than when facing an unknown length. Medium introduced this pattern in 2013 and it is now standard across most content platforms including Substack, Ghost, and most major news sites.
How reading time is calculated
The standard estimate uses 238 words per minute for average adult silent reading of non-fiction (from a 2019 meta-analysis of reading speed research). Technical content, academic text, or complex legal language slows reading to 100-150 wpm. Children read at 100-180 wpm depending on age and reading level.
Factors that affect actual reading time
- Content complexity — technical or specialised content takes longer than narrative prose
- Reader familiarity — experts read faster in their domain; novices read slower
- Screen vs print — on-screen reading is typically 20-30% slower than reading print
- Images and tables — this estimator counts words only; add approximately 12 seconds per image for viewing time
Content length and engagement metrics
Reading time estimates correlate with engagement metrics in content analytics. Articles with displayed reading times of 3-7 minutes tend to have the highest engagement rates (shares, comments, time on page). Very short content (under 1 minute) is scrolled past quickly. Very long content (15+ minutes) has high drop-off unless the topic is highly specific and the reader has strong intent.
- 3-5 minute reads — highest social sharing rate; complete enough to be valuable, short enough to finish
- 7-10 minute reads — good for evergreen how-to content and detailed tutorials
- 10-20 minute reads — long-form guides and research-heavy articles; higher bounce rate but strong for SEO
- Completion rate — scroll depth tracking shows that 55% of users spend fewer than 15 seconds on a page; reading time estimates improve this