How long should a press release actually be? At Zeest Media we reviewed more than 25,000 releases distributed between January 1 and September 30, 2026 to find out. The data shows a clear pattern: longer releases generally earned more human attention in the first 72 hours. AI engagement followed a different curve-rising at first, then flattening and declining among the longest pieces.
Human Views Rose Steadily with Length
We grouped releases into five word-count bands, from under 400 words up to 1,500 words and above. Both average and median human views increased at every step.
Releases of 1,500+ words drew 4.2 times more average human views than those under 400 words. Median views showed an even sharper gap-5.3 times higher than the shortest releases. Because both the average and the median moved in the same direction, the lift is not explained by a handful of outlier announcements. This aligns with broader patterns we have observed in our ongoing Disclosure Intelligence research.
AI Activity Followed a Different Path
AI engagement also grew with length, but the shape of the curve was different.
Average AI activity more than doubled from the under-400-word group to the 700–999-word group. It held roughly steady through the 1,000–1,499-word band, then declined among releases of 1,500 words or more.
Median AI activity kept rising across all five categories, finishing 2.3 times higher than the shortest releases. The gains became smaller as length increased, even while the average peaked and then fell.
The gap between average and median matters. A few releases with extremely high AI activity can pull the average upward. The median is less sensitive to those extremes and better reflects typical performance. This pattern is consistent with earlier findings in our report Why Some Press Releases Generate Far Higher AI Activity, which examined the wide variation in how AI systems engage with different corporate announcements. External research from sources such as the PR Newswire State of the Press Release has similarly noted that content depth often influences both human and algorithmic reach.
What Issuer Teams Should Take Away
Longer releases tend to attract more human views, yet the data does not point to a single “ideal” word count. Releases that run longer often contain more substantial news or richer context-exactly the material that readers and algorithms both value. Length itself is rarely the cause; substance is.
These findings reinforce earlier Zeest Media work showing that AI visibility depends on multiple factors working together rather than any single optimization trick. See our related analysis What Drives AI Visibility? for a deeper look at those contributing factors. Industry guidance from organizations like the Public Relations Society of America also emphasizes clarity and context over rigid length targets.
The practical goal remains the same: communicate material information clearly, accurately, and with enough context to serve both human readers and machine systems. Focus on the story first. Let the length follow the information, not the other way around.
For teams looking to improve distribution and measurement of their announcements, explore Zeest Media’s press release and newsroom solutions.

Written by
Zeest Media
Expert contributor at Zeest Media - covering PR, GEO, and digital media strategy.



