How to turn a blog post into a video
A 1,500-word article read aloud is roughly eight minutes of video. Here is how to work out which single idea in it earns the runtime, and why most lists do not survive the jump.
8 minute read
Text to video is not video to text run backwards
Almost all repurposing advice assumes you already have a recording and are slicing it. Going the other way is a harder problem, because you are not selecting from material that exists, you are generating something new from a source built for a different kind of attention. A video cannot be extracted from an article the way a clip is extracted from a podcast. What you are really doing is using the article as a research document and a structure check, then writing a script from scratch. Skip that step and treat the article as the script, and you get the thing everyone recognises within two seconds: a voice reading paragraphs over stock footage. The sentences are fine. The format is wrong, and no amount of editing rescues it.
Nobody read the article the way you think they did
Start by being honest about how the piece was actually consumed. Jakob Nielsen's analysis at Nielsen Norman Group, covering 45,237 page views, concluded that on an average visit people have time to read at most 28 percent of the words on a page, and that 20 percent is the more likely figure. Only on pages of about 111 words or fewer did readers get through roughly half the text. That is not a comment on your writing, it is how reading on a screen works. The consequence is liberating rather than depressing. The version of your article that exists in a reader's head is already heavily compressed, closer to the subheads plus the two paragraphs they stopped on than to the full eighteen hundred words. That compressed version is far closer to a video than the article is.
Do the arithmetic before you write a word of script
Work out how long the thing wants to be before you commit to anything. Marc Brysbaert's meta-analysis in the Journal of Memory and Language, pooling 190 studies and 18,573 participants, put average adult silent reading in English at 238 words a minute for non-fiction and 260 for fiction. The same paper, from a separate set of 77 studies and 5,965 participants, puts reading aloud at 183 words a minute. So a 1,500-word article delivered at a natural pace is around eight minutes of talking, before you add an opening, a pause for breath, or a single visual. That number is where the real editorial decision arrives, because you are almost certainly not making an eight-minute video of this article, which means you have to choose which one idea in it earns the runtime.
Most lists do not survive the jump
This is the part that decides whether the video works at all. A nine-item listicle is comfortable on a page because the reader can see all nine at once, skip to item six, and scroll back when they lose the thread. A listener can do none of that. Nelson Cowan's review in Behavioral and Brain Sciences argued that short-term memory holds roughly four chunks rather than the seven long assumed, and a spoken list is precisely the format that hits that ceiling. So a list either becomes one argument with two or three supporting beats, or it goes on screen as text while you talk around it. What almost never works is narrating all nine in order and expecting anyone to arrive at item nine still holding items one through eight.
The problem is structure, not speed
It is tempting to blame the format for being slow, and that turns out to be wrong. Kuperman and colleagues, writing in the Journal of Experimental Psychology: Human Perception and Performance, found that listening comprehension held up fine at 270 words a minute and only fell off sharply at 315, and that people's most efficient listening rate closely matched their own spontaneous silent reading rate for the same texts. Listening is not inherently slower than reading. What it lacks is random access, a visible structure, and the ability to reread a sentence that did not land. So when you rebuild the article, spend your effort on ordering and signposting rather than on compressing the language. Density is fine. Disorder is what loses people.
An article is random access; a video is one line
A reader arrives from search, scans the subheads, jumps to the section matching their question, and leaves. That is why good articles are built as parallel blocks that each stand alone. A video has no subheads a viewer can see before committing and no scroll bar they will use, so it needs the opposite construction: one question posed early, one path through it, one payoff. The practical translation is to pick the single most useful claim in the article, open with it or with the tension that makes it interesting, then treat every remaining section as either evidence for that claim or something to cut. If two sections feel equally important and neither supports the other, you have two videos rather than one, and forcing them together produces something with no centre.
Let the job set the length, not the word count
Wistia's State of Video report, drawn from more than 13 million videos and 79 million hours of viewing on its platform alongside a survey of over 900 professionals, found engagement rate falls as length rises: videos under a minute average around a 52 percent engagement rate, meaning viewers get roughly halfway. But the same dataset shows longer videos do better on clicks, with the 30-to-60-minute band drawing the highest click-through rate and sub-minute videos among the lowest. That contrast is worth sitting with, because it cuts against the reflex to make everything short. A one-minute video from your article is a discovery asset for people who have never heard of you. A ten-minute one is for people who already trust you. Decide which you are making first, because the two keep different material.
The video becomes a source, and the direction reverses
Once the video exists it has a second job, and this is where the whole pipeline turns around. The video you just made from an article is now a recording you can cut from. The passages that earned their place in the script are also the ones most likely to stand alone as vertical clips, and you already know which they are because you chose them once under pressure. Pulling three or four of those, reframed vertical and captioned, gives the original article a distribution route it never had on its own. That batching pass is the loop FrameOS is built for: one long recording in, a set of captioned vertical clips out in a single sitting rather than an afternoon spent scrubbing a timeline. 300 credits for 3 days · no card.
FAQ
Can I just read my blog post out loud as a video?
You can, but it rarely works. An article is written for a reader who scans and jumps around, and reading it aloud preserves the structure while removing the thing that made the structure usable. Use the article as source material, pick the one claim worth the runtime, and write a script that states it early and follows a single line through to a payoff.
How long should a video made from a blog post be?
Decide from the job rather than the word count. Wistia's data across more than 13 million videos shows engagement rate is highest under a minute, at around 52 percent, while the strongest click-through rates sit in much longer videos. A short version is a discovery asset for strangers; a long version is for an audience that already knows you. Pick one before you cut the script.
Which blog posts make the best videos?
Posts built on one arguable claim, a story, a walkthrough, or a demonstration. Posts built on a long list, a reference table, or anything a reader needs to scan and return to are the weakest candidates, because a listener cannot hold nine items in order. If you want to use a list-shaped post, put the list on screen as text and talk around it instead of narrating it.
Do I need to be on camera to turn an article into a video?
No. Screen recordings, demonstrations, and slide-led narration all work, and for how-to material they often work better than a talking head because the visual carries the explanation. What matters is that the visual changes often enough to hold attention and that the script is written for listening rather than lifted from the page.
Sources
- How Little Do Users Read? · Nielsen Norman Group
- How many words do we read per minute? A review and meta-analysis of reading rate · Journal of Memory and Language
- A lingering question addressed: Reading rate and most efficient listening rate are highly similar · Journal of Experimental Psychology: Human Perception and Performance
- The magical number 4 in short-term memory: A reconsideration of mental storage capacity · Behavioral and Brain Sciences
- How to Choose the Right Marketing Video Length for Any Goal · Wistia
Related pages
Keep reading
More Clipping & Repurposing →How to livestream and clip it afterwards
Twitch deletes most streamers' past broadcasts after seven days and Facebook deletes live video after thirty. What you record locally, and how, decides what you can still post next week.
8 minute readHow to build a clip library you can actually re-use
Four hundred exported clips in a folder is not a library. A library is one row per clip, carrying the hook line, a short tag set, and a record of everywhere it has already run.
8 minute readHow to repurpose a newsletter into video
Your newsletter is better video material than your blog, because it already has a voice and a format. Here is which issues convert, and what the Substack data says about who this actually helps.
8 minute readTurn one long video into a week of posts
300 credits for 3 days · no card.
Start free