Infinite Scroll for News Sites

I’ve never been a big fan of the infinite scrolling story pattern because for some reason, for me, it does the opposite of what it’s supposed to do—it adds friction. I’ve always chalked it up to personal preference but I was curious to see if I could find an objective argument.

Premise: the web is a collection of resources—sites, pages, files—and the way we get from one resource to the next, that is, the way we request a resource, is by clicking.

Once we receive the resource, the way we maneuver within it is by scrolling. Scrolling is an interior action for activity sandboxed to the current context.

We can easily visualize this by imagining clicks moving us from resource to resource along the X axis, and scrolling moving us up and down within each resource along the Y axis.

Of course, clicks are also used to navigate within a resource. This is OK because the click is the dominant of the two actions. It has all the privileges of the scroll plus additional ones.

The limited permissions of scrolling is part of our contract with the UI all the way down to the OS level. We can scroll up and down the iTunes Terms & Conditions but only a click will give our consent and get us to the next screen. On Amazon, scrolling to the bottom of the product details does not automatically add the product to our cart. You can’t reboot by scrolling.

So another way to look at the click is that it’s used to handle the important stuff, the stuff that requires some level of authorization.

Now we get to the problem with infinite article scrolling. Scrolling doesn’t have permissions to request new resources! It’s a violation of the contract. But only if you consider an article to be a resource.

So to the extent that this idea of scroll authority vs click authority holds water, the question is to what extent do we expect articles to behave like resources?

I think of an article as an encapsulated, complex entity, with lots of auxiliary information and actions attached. It’s too big to hold in my head as a list item. Stories have a beginning and an end and a reflection of that in the medium works as an affordance. From the user’s perspective, the seams between articles don’t create friction—they help keep everything in order.

So when these divisions become blurred, it can create a feeling of uneasiness. It can create friction.

But this is speculative and subjective (for some of the more standard UX arguments about infinite scroll, see here).

The Publisher’s Perspective

To the publisher, the seams between stories do present a form of friction by requiring the user to take action to continue spending time on the site. With the infinite scrolling story design pattern, the publisher trades the higher interaction cost of the click for the lower interaction cost of the scroll in an attempt to leapfrog the “next action” decision. Instead of waiting for the reader to choose another story, one is chosen for them and the reader is immediately transitioned into it.

This is interesting in a few ways but one thing I’m curious about is the question of what happens to the perceived value of the next story when we take away the reader’s opportunity to choose it. Selection imparts value. I’m more likely to value a story I’ve chosen (even if it was suggested to me), which means I’m more likely to pay attention to it, more likely to share it, and so on. Maybe.

I’d bet that scrolled-to stories have lower engagement rates and lower share rates than clicked-to stories.

Of course it may be hard to measure: scrolling sites encourage passive scrolling which devalues scrolling as a reliable engagement metric. Which gets into a whole other topic of how the continuous scrolling stories pattern disrupts the old-model metrics.