Infinite scrolling essentially allows a user to scroll endlessly (until there is no more data to fetch) through chunked data. It is usually implemented for comments, blog posts, product listings and other “chunk able data”.
Such an implementation requires the pre-fetching of content from a subsequent page and adding it to the user’s current page.
Seems like a yay for most things — i mean who wouldn’t like to click less next/more result buttons?
But the implementation at Etsy seemed to say otherwise. Read the article here, which discusses plausible reasons including
(1) Some users, when returning to the search page after backing up from a product page, would find that they had lost their place in the infinite search stream (probably the biggest problem with infinite scroll implementations). Hence, it seems like without typical pagination, users typical navigational patterns when thrown askew.
[It then delves into an interesting issue of Etsy assuming a typical power-user behaviour (of open new tabs more each new product search) as the default behaviour.. but that’s another topic for discussion in the future.]
(2) Was it psychological? Perhaps users simply did not want to feel like they were facing with an overwhelming number of choices. There was perhaps an illusion of limited choice with pagination, where perhaps the additional action of clicking the next button acted as a psychological breakpoint in limiting a user’s possible choices. With infinite scrolling, this psychological breakpoint disappeared and users ended up being fatigued by the choices.
It’s so interesting that such psychological reasons can have such a great impact. But what the Etsy experience tells us is this – there is value is quick, dirty and modular testing — user interactions are becoming such a nuanced feel that any change or addition could have unexpected and undesirable consequences.
Don’t assume just because infinite scrolling happens to be the next big design trend that it is the next feature one should pursue.