To Like or Not to Like: Will Killing the “Like” Button Save Insta?

By Victoria Mariolle

On July 17, Instagram announced the unthinkable: the company was exploring the idea of hiding the number of “likes” from its photo-sharing platform. According to the company, the new design would encourage “followers to focus on the photos and videos you share, not how many likes they get,” ostensibly shifting the emphasis away from the quantity of likes to the quality of content.

The news brought a wave of mixed reactions. Social media influencers grieved at the possibility that the change would undermine their online popularity—and profit margins. Melbourne influencer Mikaela Testa sobbed, lamenting that her “blood, sweat and tears” were all for naught. Fitness and food influencer Jem Wolfie told Hack, “They’re taking a tool away that’s really important for us. I’m still going to keep posting as much as before, but it’s demotivating for me.”

While influencers fretted about their social worth, high-ups at Instagram have championed these changes as an important step for users’ mental and physical health. A 2017 report published by the Royal Society for Public Health, an independent public health charity, found that out of all social media platforms, Instagram was ranked the worst for adolescents’ emotional wellbeing. (Participants reported getting less sleep, feeling worse about their bodies, and experiencing greater “FOMO,” or fear of missing out.)

As beta-testing for this new, Like-less version of the app is underway worldwide—just last month, Instagrammers in the U.S. became the newest guinea pigs—the question remains: Can tweaks in the user-interface design really save us from our worst online tendencies? (Interestingly, Facebook, owner of the photo-sharing platform, recently announced that they too are considering hiding the like count—perhaps, some speculate, to conceal their shrinking popularity.) We spoke with Galen Panger, a former PhD at UC Berkeley’s School of Information and a user design expert at one of the Big Five tech companies, to find out. This conversation has been edited for length and clarity.

Why would Instagram implement this change now?

Instagram could be trying to address some of the negative externalities of their service. There is a broader social criticism out there of all these companies—Twitter, Facebook, Instagram, etc—for using metrics to measure people’s worth. Twitter has recently reduced the font size on the follower count to, in at least that extremely minor way, de-emphasize the number of followers you have.

The other thing I think is really important for Instagram is how much people post. What does Instagram want the platform to be? Do they want it to be a place where people feel more comfortable being wacky and less varnished? Maybe some people feel like all they ever get likes on are the selfies or the vacation photos they post. So hiding the likes means that, if you do experiment more, if you do use Instagram with a little bit more whimsy, you don’t feel like you have to delete those posts if they don’t get the likes that you expect.

What do you think of the backlash Instagram has received from influencers and general users?

As far as I’m aware, people will [still] be able to see the number of likes that they’ve gotten personally on each post. I imagine there’s some way for these influencers to screenshot that, or share their account with a brand. I’m sure that Instagram really cares about these influencers because it’s a big part of the engagement on that platform. I think that it’s valid for [users] to want that kind of validation and to publicly show how much validation they get.

[But] the majority of people aren’t getting tons of likes. And there is the argument that likes can be a source of envy because of something called social comparison. In the absence of more objective measures, we use substitute measures, like “likes” on friends’ posts, to see if we measure up. There’s a long line of research—long for social media—showing how people develop a sense of inferiority seeing the polished lives of their friends online. That doesn’t really match the day-to-day normalcy of life.

Do you think the single option to “like,” or not, on Instagram as opposed to the many reactions available on Facebook, makes the user experience on Instagram more negative?

These reactions have always been difficult because it’s a one-button signal. Originally Facebook was just a “like” as well—either you’re getting a “like,” or you’re having an absence of “like.” There could be different reasons for an absence of like: nobody saw the post because Facebook’s algorithm has determined no one wants to see it, or it didn’t catch anyone’s eye and everyone just kind of scrolled past it.

Even when you expand to five reactions, that doesn’t solve that problem. But it does help solve the problem of the “like” button being an inappropriate response. When someone posts news like, “President announces policy to separate children from their parents at the border,” it’s really awkward to like it. Saying you feel angry seems more appropriate. If someone says, “My grandfather died,” liking that post is very awkward, but saying you feel sad makes more sense.

How do likes affect Facebook users’ behavior compared to Instagram users?

Facebook is a little bit lower-pressure. You can be a little more free-wheeling and post random thoughts. Most of the research that we’re applying to Instagram originated from Facebook—where people tend to self-censor. There was one study that was done by Facebook that suggested one out of every three posts is censored at the last minute. People start to compose a post, and then they delete it. So you have a really high level of self-censorship. That may apply even more strongly to Instagram.

In 2017, Instagram changed its feed from displaying content chronologically to algorithmically. What effect do you think this change has had on the content we see?

There have been some studies on whether the “filter bubble effect” is something that’s really substantial. But the filter bubble tends to be more a result of who we follow and connect with, than it does the algorithm showing us things that fit with a particular viewpoint. The algorithm may contribute some, but the biggest issue is whether or not you follow a diversity of people across your interest areas. Political parties, Democrats and Republicans are geographically separated, right? People are filtering themselves and sorting themselves geographically. It’s not just happening online. This is something broader socially than just the algorithms.

How does this new algorithmic feed affect users who engage more or less often?

With reverse chronology, your feed was dominated by the people who posted the most. With more of an algorithmic style, you might actually get to hear from those people that you don’t hear from as much. Now, it’s still going to try to show you people that you frequently interact with. So that is one bias of it. But it’s also maybe a bias that makes the experience a bit better.

Instagram has rolled out something that I think is really cool, which is the “you’re all caught up” feature. It tells you that you can stop using Instagram now because you’ve seen everything that people have posted in the past couple of days, even from those friends who we might miss out on because you don’t interact with their stuff, or they’re less verbose. The algorithmic changes really have costs and benefits. I just think your perspective on it depends on which things you value the most.

Share this article:
Google+ Reddit

Add new comment