Here’s a question. Today, Instagram/Threads chief Adam Mosseri posted this:
Okay, all good, that makes sense, as engagement is more important than follower counts, which can be gamed, cheated, etc.
But that being the case, why display follower accounts so prominently in both of your own apps?
Like, you’re the one in charge, buddy, why are you making these metrics a focus, by making them so “prominent and easy to find”?
You can change it if you want.
Over time, this has repeatedly been identified as a potential issue for social media apps, as they examine the incentives that drive negative behaviors and experiences.
Former Twitter chief Jack Dorsey, for example, noted back in 2018 that by emphasizing follower counts, that then drives individuals to post more polarizing content, as that then boosts their exposure potential, and helps them gain more followers. And the apps themselves are inadvertently incentivizing this.
As per Dorsey:
“[We initially] made the [followers] font size a little bit bigger than everything else on the page. We didn’t really think much about it, and we moved on to the next problem to solve. What that has done is we put all the emphasis, not intending to, on that number of how many people follow me. So if that number is big and bold, what do people want to do with it? They want to make it go up.”
Twitter sought to address this by… making the follower count font slightly smaller.
Which, as you can probably guess, was not an overly effective approach.
Instagram, too, has made moves to redirect user incentives, by hiding like counts on posts, in order to “depressurize Instagram for young people”, according to Mosseri himself.
So Mosseri, and Meta, are well-versed in the potential harms and negative behaviors that focus metrics can drive.
Yet, when designing Threads, what got prime placement?
I mean, you could argue that Threads has made the font here lighter, which reduces its presence slightly. But again, if Mosseri believes that follower counts are not a great indicator of value, and engagement per post is more relevant, why not list that instead? Why not display an average likes/comments per posts ratio or something in its place?
You can do this, Mosseri. Nobody else can. So why are you telling us?
This is the latest of Mosseri’s various stances on social platform engagement which seem to contradict the norm, yet continue to be implemented in his own apps.
Back in January, Mosseri said that things like Trending Topics won’t be overly valuable for the app, at least not to the degree that many expect.
As per Mosseri:
“My honest take is that requested features like lists, an edit button, a following feed, trending, and hashtags are all good to build, but none noticeably grow Threads or Threads usage. We’ll continue to build them because it’s good to build features that your most engaged users are excited about, but it’s hard to prioritize them when the measurable impact is negligible.”
Yet, Threads has implemented, or is building (as Mosseri notes) versions of all of these.
So why? If Mosseri knows that these are not highly valuable, and he has alternative approaches, and it’s worth noting here that Adam Mosseri has been leading engagement projects at Facebook and Instagram for many years, so he probably would know what works.
If he wants to take a different approach, why doesn’t he? Why tell us how to use the app, in variance to its UI, if you have a better option?
And there may well be value in a different approach, in a different way of looking at things. Maybe removing follower counts and follower lists would help to reduce the emphasis on this element. You have an opportunity in Threads to change this paradigm, why not try it out?
It just seems like an odd stance to take, when you’re the one making the calls on such.
I do think that there could be merit to making such changes, in order to see if/how they impact usage behaviors. But it’s all academic till someone takes the leap and tries it in practice.
And the person in charge of an evolving social platform is the one raising it. Seems like a good time to test.
Source link
