The End of Pull-To-Refresh

Loren Brichter integrated the scroll and refresh events into one single action, and so pull-to-refresh was born. At the time, pull-to-refresh was regarded as an innovation, but has it lost its purpose? and worse, has it inadvertently produced a nasty side-effect that could change Twitter forever?

Personally, I never quite liked pull-to-refresh. It makes me feel uneasy. Just look at the top of this post, a tension is created. A slot machine of sorts, will I or won’t I have something to look at? or worse, feel good about. “Loading…” Just one more second. It silly to me that we ever used it, but more so that we continue to – especially given the efficiency of instant messaging. There’s a reason why we don’t pull-to-refresh in WhatsApp or iMessage.

Pull to refresh is a hack. I never liked it as an interaction. pic.twitter.com/U9lLV34uub

— Youssef (@YS) December 30, 2013

Twitter recently made a change to the feed algorithm which now injects related Tweets into your main feed. Tweets that people you follow favorite are now frequently promoted into your main feed. This version of Twitter feels less like Twitter and more like Facebook. It no longer feels real-time, and here is how it comes full-circle back to pull-to-refresh. I asked Dick Costolo, Twitter’s CEO, why they made these changes to the feed and here was is explanation:

@YS @mjbiren you get favorites when you pull-to-refresh 2x and we have no new tweets from follows both times.

— dick costolo (@dickc) August 31, 2014

@YS @mjbiren … 2nd request for more content that would otherwise come back empty. Fulfilling that request well is a worthy goal, I think

— dick costolo (@dickc) August 31, 2014

One can assume that Twitter saw a lot of data which indicated people were refreshing their feeds, only to be disappointed with no new Tweets. So, rather than fixing the problem by removing pull-to-refresh, they are giving us a bunch of random Tweets to justify the refresh.

This is a classic case of treating the symptom, rather than the cause, which is a common mistake many design teams make. I feel it’s time we retire pull-to-refresh, so we can get on with the real business of using Twitter, and not its nostalgic interface.