Skimming and skipping is not a problem; quit trying to fix it

On a quite regular basis, someone publishes a study proclaiming — horror of horrors — that on the Web, people don’t always read content all the way through but skim and skip. Oh no! The Web is rotting our brains! Oh no! Writers have to change everything they do!

Bosh. Did they not bother to ask if people skipped and skimmed before the Web? Did they ever pause to consider how people read newspapers or the magazines in doctor’s waiting rooms?

skippingSkimming and skipping is perfectly normal reader behavior. It does not indicate that there is a problem with the content or with the reader. It is not a problem to be fixed, though it may be something that writers need to do a little bit more to support.

This rant is provoked by this LinkedIn discussion, which points to an article that, essentially, redubs progressive disclosure as “hypotext”.

Several folks (including myself) commented that the kind of information hiding it describes is a bad idea. But here I want to take a step back from that and look at the underlying notion that reader’s skipping and skimming behavior is a problem that needs to be fixed, rather than totally expected and appropriate reader behavior.

The underlying idea here seems to be that if a piece of content is good, then every reader will want, need, and appreciate, every single word of it, while requiring no other words at all from any other piece of content. But that expectation is utterly ridiculous.

If you paved a path, would you regard it as a failure if someone did not step on every paving stone, or if they ever wanted to go somewhere not on your path?

Maybe it is the writer’s ego. We want every word to be consumed and appreciated. Maybe it is the influence of the school system which still puts so much emphasis on learning texts and passing comprehension tests, rather than actually doing useful things in the world. Maybe it is a genuine zeal for the reader’s convenience which makes us think that if they are skimming and skipping, they must be unhappy.

In the LinkedIn discussion, Fer O’Neil describes an experiment in which his group used dropdowns to hide screenshots from a procedure and tested it with a group of users. Predictably, they hated it.

Why predictably? Because we are really good at skimming and skipping. It is how we read and it takes no learning and very little mental energy. It is by far the easiest way to move through a text of any reasonable size. Figuring out that some text is hidden and that there is a mechanism for revealing it, on the other hand, takes learning and mental energy and interrupts the reader’s flow and concentration.

But my point here is not that the solution is bad, but that the problem is non-existent. Skipping and skimming is normal reader behavior. It does not need to be fixed, and readers will not thank you for the attempt.

Every reader is on an individual journey, trying to find the information they need to gain the confidence to attempt some task. Exactly what they need to get to that point varies based on many variables.

  • What exactly is the task.
  • In what context, and with what constraints, is the task performed.
  • What similar tasks has the reader done in the past.
  • How well does the reader understand the background of the task and the tools.
  • How well does the reader’s education and experience equip them to understand the language in which the content they are reading is expressed.
  • How confident is the reader generally, and in this type of task.
  • How motivated are they; what is at stake?
  • How urgent is the task; do they have time to wait or must they act immediately.

With so many variables in play, every reader’s journey is unique. And because these factors can change from day to day and hour to hour, the full journey cannot be anticipated even for a single reader attempting a single task. The reader could not anticipate it, let alone the writer.

All content, therefore, is written for a reader who is on an individual journey. Just as a traveller drives an individual route in a generic car over generic roads, the reader’s journey is made up of generic content, written to serve the unknown goals of many readers. Having to navigate, steer, and make decisions are part of the necessities of travel. Having to search, link, skim, and skip are part of the necessities of research. Travellers and readers have fully accepted them, and mastered them. Good content, like a good road, is easier to navigate, but it does not obviate the need for navigation.

Any attempt to meet these needs in a more direct way would require dynamic real-time information about the exact state of the reader’s learning as well as the ability to intuit which piece of information would exactly meet their next need. We may know more about our readers than ever before, and may even be able to tailor content to their needs to a certain extent, but we will never have the data, nor the algorithm, to roll out the perfect text before their eyes in real time.

Anything less than this, though, simply disrupts the reader’s well established and highly efficient skimming and skipping behavior which they use to navigate most texts that they read. Content navigation is not improved by fancy do-dads that try to eliminate skipping and skimming, but by well structured text that supports them.

That readers are skipping and skimming through your content is not, in itself, a problem. It is the behavior we should expect and support.

We need to stop trying to fix things that are not broken.

There are quite a few things that genuinely are broken. We should focus on those.



8 Responses to Skimming and skipping is not a problem; quit trying to fix it

  1. Stephanie 2015/07/07 at 11:21 #

    Good read! I actually skimmed most of this and then went back and reread it in more detail just out of habit. Ironically, I like to write but do not like to read a lot.

    I think if a document, website, or whatever you are reading has good structure and consistency to it then the reader can easily identify the main points that stand out. For example, I write procedures and the style I use is to bold action items and the screenshots below the step typically provide a call out of the action item. The user wouldn’t necesarilly have to read the procedure word for word to find what he/she needs but would know to look for the bold action items in a step to know what to click or look for the call out in the image to know what to look for on the screen.

    • Mark Baker 2015/07/07 at 11:39 #

      Thanks for the comment, Stephanie.

      I think that is a very common pattern — skimming and then going back to reread if your interest is sufficiently engaged. I’m glad this post passed that test for you!

  2. Leigh White 2015/07/07 at 11:28 #

    Wow, Mark. For once, I almost completely agree with you 🙂 I don’t think progressive disclosure is necessarily a bad thing or that it requires more mental effort. In a way, any kind of hyperlink is a kind of progressive disclosure. We figure out quickly enough that it’s a portal to additional information that we can choose to see or not see. Other than that one point, you’re right on. We should embrace skipping and skimming. Our responsibility is to try to anticipate the most salient points and do what we can to ensure the reader’s eye falls easily upon them. It’s like landscape architecture on the page.

    • Mark Baker 2015/07/07 at 11:49 #

      Thanks for the comment, Leigh.

      I don’t think progressive disclosure is always bad either. I’m pretty sure it is always more mental effort than scanning, because the mechanism is not universal and you have to select something to click on, but there can be cases where that additional mental effort is worth it to save time, for instance. Most of the time, no, and it does not save any time. But sometimes yes.

      And I agree too that linking can be considered a form of progressive disclosure. But linking is (or should be) to ancillary or related material, not to parts of the structure of the current topic. (Reuse mechanisms sometimes mix up these categories, so that can be a problem.)

      So linking is not really about progressive disclosure of the current, but onward navigation to the next. Certainly, though, I see examples of progressive disclosure that should be links to separate topics. Indeed, the reuse mindset can encourage inlining material where it should really be linked to, which then creates the need to hide, since it does not really belong inline in the first place. This is why I think it is important to encourage hypertext thinking, and to focus on reduce before reuse, as discussed in my last post.

      • Marc Gravez 2015/07/29 at 12:21 #

        Mark, I think where this going is that, like so many other things related to communication, progressive disclosure is effective when used appropriately.

        • Mark Baker 2015/07/31 at 15:41 #

          Thanks for the comment, Marc.

          Agreed. There are times when it is appropriate. FAQs are an example. Hiding the answers makes it easier to scan the questions.

  3. Marie-L. Flacke 2015/07/08 at 03:53 #

    What about Gerry McGovern’s “Top Tasks”?


    (b) Top Tasks Identification Overview, webinar, July 8

    • Mark Baker 2015/07/31 at 15:49 #

      Thanks for the comment, Marie-L.

      I think when you can identify certain tasks as more common than others, you should certainly try to make them the easiest to find. Even when the reader has found their task, however, they may skim and skip around trying to fill in the gaps in their knowledge or check for an alternate solution. In other words, even if you identify your reader’s top tasks and put up big signs pointing to them, that does not mean the reader’s journey ends there.