Deutschland - de
Wählen Sie Ihren bevorzugten Standort

The power of likes on social media: Friend or foe?

These days, like buttons are part of the internet furniture. Everywhere from social media, video platforms, news outlets, and e-commerce websites make use of like buttons to allow users to signal how they feel about a post, some content, or even a product.

Whether on Facebook, Instagram, LinkedIn, or even Amazon, we’ve all probably clicked a form of like button at some point. They allow us to signal to others that we think a post, content, review, or product is useful or relevant to us, or just that we simply like it.

Likes on social media are a form of communication allowing us to signal our validation and approval with a single click, without having to type anything. What may seem like a fairly simple and basic feature is actually one the most powerful online tools ever created.

The history of the like

Likes were first introduced in 2005 by the video site Vimeo, but they really came into the wider public view when they were adopted by Facebook in 2009. At first, Mark Zuckerberg wasn’t keen on the idea, but eventually caved to pressure from his team, who were excited about the initial ‘Awesome button’.

Their reasoning: many posts had a lot of duplicate comments (‘Congratulations’ ‘Well done,’ etc.), and a simple like button would make it easier for users to signal approval and reduce duplicate comments.

It would also allow users to very quickly understand how popular or relevant a post is without having to read all the comments. From that relatively simple beginning, the like button quickly became a phenomenon. Content on Facebook is now liked more than 3 billion times per day and estimates suggest that since its inception, the like button has been pressed many trillions of times.

This is just on Facebook alone. The like button is now everywhere, from Instagram to LinkedIn, as well as blogs and news websites. Whenever a platform wants to give users the ability to quickly and easily express their opinion on something, the like button (or a form of it) is there.

The power of data combined with the power of likes: Disquieting or delightful?

I’m pretty sure that when the like button was invented, no one quite realized how powerful the data it creates would be. It’s recently been said that data is now the most valuable commodity in the world, and the like button creates vast amounts of it. If we see a post or video on Facebook, Instagram, or LinkedIn that we find interesting, we like it. It’s quick and easy; I doubt many of us are considering what that data is creating about us.

To highlight this, in 2015, researchers from the University of Cambridge and Stanford University released a study illustrating how Facebook like data can be used to accurately predict a user’s personality traits in a very powerful way. Back in 2007, researchers created an online personality test and posted it on Facebook. To take the test, users needed to give the researchers access to their Facebook data.

The test went viral and over 80,000 people took the test and provided access. This meant researchers had not only the results of the personality tests, but also access to a vast amount of Facebook data. The thing that everyone forgot is that in giving a third party access to your data, they weren’t just giving them a snapshot of the data, they were granting them access to future data as well. Facebook has since updated their privacy policies in light of recent scandals, and it’s now much harder for apps to get, and retain, access to personal data.

Using the data collected from the personality test and the Facebook like data, the researchers created a computer algorithm to understand how accurately they could determine a user’s personality profile based simply on Facebook likes.

They wanted to test whether Facebook like data could more accurately predict personality traits than humans. Incredibly, the study showed that with a remarkably small number of likes, the algorithm could determine the user’s personality to a very high degree, AND out-perform people who know the test subject personally.

The study eventually concluded that with just 10 likes, the algorithm would know you better than a work colleague. With 150 likes, it would know you better than members of your immediate family, and with just 300 likes, it would know you better than your spouse.

Just think about that for a moment: If you’ve liked more than 300 posts since 2009, Facebook probably knows you better than anyone; even your spouse. That’s quite an incredible – and scary – thought.

You can easily imagine how this kind of data and insight could be very valuable to companies when trying to provide customers with relevant and personal experiences.

Take an example of a ski equipment store opening in your area: if you’ve previously liked videos and posts related to skiing and Facebook knows where you’re geographically located, it makes sense to display a post advertising the new store to you rather than someone who has never shown an interest in skiing or doesn’t live in your area.

The ski store avoids wasting money advertising to people who don’t find them relevant, and uninteresting ads are removed for users. I don’t think anyone would have too many complaints about that. Facebook can go a lot further, and probably knows what kind of skiing you like, where and when you like to ski, and even how good you are.

It can now start to show you ads and content that’s much more relevant to you, which sounds like a win-win situation. Most e-commerce businesses attempt to segment their customers based on the data they hold about them, but this is usually quite basic. They may segment on gender, geographical location, or even what the customer has previously bought. This is obviously a very good start, but do they really know the customer? Do they understand their personality? Do they know them better than their spouse?

The answer is obviously no, but this illustrates how powerful like data is. Maybe e-commerce companies should provide more content to customers, create a community, then allow customers to like posts or content. That would enable them to start to really knowing the personality of their customers.

The darker side of likes

While it’s ideal to think that this data is most often used in a beneficial way, it’s not too much of a leap to see how it can used for much more sinister purposes.

If someone is of voting age, lives in a particular area, and has previously liked posts that criticize a particular section of society, and is also friends with someone else who has expressed certain views, it’s not hard to see how this information could be used to target them with political or inflammatory content.

In fact, it’s been claimed that this kind insight was used to influence both the 2016 US Presidential election, as well as the UK’s Brexit referendum, through a company called Cambridge Analytica.

What’s to stop foreign powers using this data to influence elections, or even incite political unrest or even violence? If Facebook knows your personality better than your spouse, it can’t be too hard to use that data to influence you in a certain way. You may have taken personality tests on Facebook; finding out which Game of Thrones character you are most like, but did you consider who is behind the test and who you may be giving your data to, and what that data may be used for?

There is another, less appreciated, but worrying aspect to likes: validation, addiction, and how it can affect mental health. Likes, and the data behind them, can be very powerful tools to allow companies to provide customers with a better customer experience but, on a personal level, there’s plenty of evidence that likes, or the drive to attain them, can be incredibly damaging for some individuals’ mental health.

Likes are a basic, yet powerful, form of validation. I suspect most of us have been a little disappointed when a post we’ve made gains very few likes, even though we thought it would get more. For most of us, this is probably a minor disappointment, but for some it can cause a great deal of anxiety.

A article published in July 2017 contained an interview with Leah Pearlman, who previously worked at Facebook, and was a member of the team that originally came up with the concept of the Facebook like. In the article, she describes how she used to draw comics in her own time and would post them on the platform, getting quite a few likes.

She then explained that in 2015, Facebook changed an algorithm, which meant that less people saw the comic strips and she received a lot less likes. Even though she consciously knew this was due to the algorithm change, she found it so hard to deal with the decreased numbers that she started to buy paid ads to ensure the comics were shown to enough people that she’d receive enough validation. This is someone who was intimately involved in the creation of the Facebook like button, but resorted to paying Facebook to ensure she got enough likes on personal posts. Likes had become a form of validation for her, and this validation became so important to her that she paid real money to get it.

It seems this is something Facebook is beginning to understand – they very recently announced they were trialing hiding the likes counter on Instagram, which the photo-sharing app VSCO has been doing since their debut. The creators of the VSCO app said that hiding all comments and likes spurs user creativity by removing the pressure of appearing popular, so that they can focus on the art instead.

At present, the initiative by Facebook is designed to stop user envy surrounding the number of likes on other posts. Like VSCO, I assume users will still be able to see how many likes their own posts have received, but at least they won’t be comparing it to the number of likes others get. It’s a start and recognition that likes aren’t necessarily as healthy for users as they may appear to be.

The future of likes is uncertain. One one hand, they’re an extremely powerful and valuable tool for platforms such as Facebook, Twitter, and Instagram, and are used to fuel billions of dollars of advertising revenue. They help to provide users with rich and relevant content, and can give us a much better customer experience.

On the other hand, if this data is not carefully governed, its use can extend far beyond providing better customer experiences, and step into a world where it’s manipulating society and causing harm to users. Either way, it’s one of the most powerful tools ever created, and is unlikely to be going anywhere soon.


This article was first published here:  

For more articles by Branwell Moffat also see