Reading Time Estimates

When I implemented reading time estimates on this site, (for Squarespace and then for Statamic) I did it because I could. I didn’t put a lot of thought into whether it was a smart move, I just noticed the feature on a few sites and thought: there’s an interesting problem I could solve.

“Ah,” I hear the snarky amongst you cry, “you shouldn’t just do things because you can.”

Thing is: I’m not selling software to thousands of people and reporting back to investors, here. If I want to spend half an hour making an unnecessary feature for my site, that’s my prerogative.

That said, I decided I’d evaluate after a while. Is there actually any value in a reading time estimate?

Arguments against

The strongest case I’ve heard in opposition is twofold:

  1. Reading time is a poor metric to judge a piece by. It gives the impression that the most important aspect of a wad of writing is how long it’ll take you to munch through it, not the quality, topic, or anything else. It enables the “long-form fetish”.
  2. Reading time indicators are chronically imprecise. People read at widely varied speeds. Whatever number you estimate is going to be wrong for almost everyone.

Ok, but…

I agree with these arguments. That a piece is long tells you nothing about it, qualitatively. As my wife frequently reminds me when we’re driving: the only thing someones flashing indicator means is that their indicator bulb is working. The only thing a long word count means is that the author’s keyboard works.

Reading time estimates are imprecise. In fact, I think they’re so imprecise as to be completely wrong in almost every case. There are far too many unknown factors. That said, I think that some indication of the length of a piece can be helpful, as long as it’s done right.

When you pick up a book you can tell roughly how long it is by the heft of it. When you start a magazine article you can judge it by pages or column inches. Reading things on a screen presents a new challenge. The content isn’t bound by the physical constraints that we’ve internalized over decades of interacting with more physical media.

Not everyone knows the “check the scrollbar” trick. If you start reading something and you’re enjoying it, sometimes it’s nice to have an indication that it’s going to take longer than a couple of minutes, so you can save it for later. Conversely, it’s nice to know that you can probably finish it in a short amount of time.


I’m keeping the length indicators, because:

  1. Some readers have told me they like them.
  2. I very much doubt that they’re obtrusive enough to put anyone else off.
  3. If anyone is judging my articles by how long they are, I really don’t care. That’s their deal, not mine.

However, I do get annoyed by the imprecision of x min read. So instead of providing a number count, what I’ve done is changed the script to display a different phrase based on the length of the article. Hopefully this will be mildly informative and maybe even a little entertaining.

(If you’re really interested in knowing what all the phrases are you can view source. I know that’s what I’d do.)

EDIT: For now, I’m just displaying the number of pages the article is on the reader’s current screen.

Related Entries

Author Chunking