First Paper Out: a Critical Analysis of (Aesthetic) Design Power

I’m excited that my first-authored paper, “The Shape of Social Media: Towards Addressing (Aesthetic) Design Power“, has just been published at DIS2022!

While we know a lot about the effects social media have on us as individuals and as a society, we don’t really know how this connects to specific design features yet. There’s a bit of work that speculates about design, or proposes design interventions, but not much.

So I did a critical analysis of four social media platforms that are currently very popular in Austria – Instagram, TikTok, Twitter and Facebook – to establish and provide a foundation for critical discussions around how design, in this case particularly formal and aesthetic design, could be connected to their effects.

Here’s the presentation video:

(featuring an unfortunately conspiration-theory-esque still):

And here’s a short summary of the paper:

Chris (my supervisor) and I coined the term “Design Power” as an umbrella term to encompass existing theories (e.g. persuasive design, nudge, dark patterns, and advertising). Social media and technologies permeate all aspects of our lives, so we need to adjust our vocabulary to encompass not only the multi-faceted ways that power manifests in design, but also the moral ambiguity (Design Power is not evil by nature!) and the role and agency of users (Design Power can be negotiated, redistributed and regulated).

Owing especially to how the field of HCI really likes to focus on function rather than form, in my analysis, I focus (as much as possible, of course – form/function is a duality after all) only on the Aesthetic Design Power.

Here’s some of the main findings of the analysis:

Social media are increasingly homogenising in terms of aesthetic design.

Many icons are, if not almost exactly identical, created in the same style, and the layouts and menus are extremely similar. Platforms assume that users are familiar with icons and functionalities. Interface elements are generally as “minimal” and inconspicuous as possible. Identity is marked uniformly as round visual avatars, only sometimes accompanied by usernames.

Interactions are highly and similarly mediated

Both content and framing across all four platforms is rigid, uniform and uncustomisable, with a high level of design mediation. For instance,
platforms either only included possibilities for reacting to content positively, or of choosing from 7 pre-defned, animated emoji for their reaction. The mediation in public areas is higher than in private chats.

Social context is emphasised, using similar visual language

The exception to this backgrounding of visual design elements are any social features and notifications, which are high-contrast (e.g. likes, hearts, new messages), always colourful and frequently animated when tapped. Everything else being so minimal makes these elements stand out all the more. Red is frequently used as a signal colour, as is green.

Ephemeral content is SIMILARLY EMPHASISED

Ephemeral content is always prominently placed and visible first thing at the top of the screen (where applicable), marked by strong and flashy colours. Once viewed, the colourful elements usually grey out. One of the newest features, live video sharing, is placed the most prominently and aggressively visually highlighted. For TikTok, it simply is the landing page.

TikTok: Getting rid of negative space altogether

TikTok is not just the newest of the social media analysed, it also has the youngest user demographic. You immediately find yourself watching an endless stream of content filling up your entire screen. Interface elements are overlayed, like a video game HUD, making TikTok extremely immersive. Content autoplays, so there is also a lack of negative temporal space. Similar new types of content on other platforms have recently been introduced.

When DESIGN principles are applied to hide

Often, instead of analysing when design is used to signalise, it is more interesting to see in which cases design is used to hide and obscure. This is the case on all platforms when it comes to advertisements: not only are advertisements visually almost identical to regular user content across all platforms, but words like “promoted” “sponsored” or “paid partnership” are used to mark them. These tags are never visual, always verbal, low in contrast, small, and grouped with other metadata.

It is also a mystery why and which certain few comments are displayed under content. Are they the most interacted with? The most polarising? The element you have to tap to access more comments is also decidedly inconspicuous compared to tokenised heart or smiley reaction possibilities: it is greyed out. Responding to content freely is not encouraged.

What Does this all mean?

If we assume that design serves the (economic) goals of social media companies, then currently, aesthetic design choices are made to promote ephemeral content, deep immersion, socio-emotional contextualisation of content and superficial social interactions.

Social media are not just homogenising, but becoming richer and more similar in features. While at first glance, this may seem to empower users and shift responsibility towards users in terms of how they use social media, the way
these features are visually presented and ordered contradicts this. Existing research also indicates that people with pre-existing mental health conditions are more prone to have unhealthy use patterns and be more vulnerable to the negative effects of social media.

Higher mediation of social interactions means higher vulnerability to Design Power, especially in public interactions. Previous research shows that directly interacting with people close to us makes us happier, while passive interactions with strangers, brands and celebrities promote upward social comparison and make us unhappier.

But simply not using social media can have other negative effects, and for many, logging off is not an option. For instance, social media provide an important resource for many during the current COVID-19 pandemic, and continue to be important spaces for marginalised groups to build communities.

We can speculate about some of our findings, but on others, we are in the dark. Regarding what choosing from only the same 8 emoji to react to any information we receive does with us, how much work it takes to recognise advertisements, and how the inconvenience of one extra tap scales to have larger effects: we simply do not know enough about the design of social media to make any claims, and we desperately need more research!

We need to talk about regulating design

Explorations of alternative designs reveal and challenge Design Power. SFor instance, Not for You by Ben Grosser confuses TikTok algorithms, while Lyngs et al. present design variations of Facebook against distraction. Both academic and public voices call for more user control, power and responsibility.

Recently, activists and policymakers begun to engage the importance of design. In 2021, privacy activists NOYB sent more than 500 complaints to companies about their cookie banners, whose misleading design is in violation of the GDPR, while in 2022, the EU reports on dark patterns and user manipulation in their study on unfair digital commercial practices, connecting physical and psychological reactions to specific dark patterns and point towards the structural harms and inefficiency of current countermeasures.

Companies can and should implement more positive design. Social media companies are more than capable of changing their design if there is enough public pressure. But waiting for public outcry to create enough pressure to change salient “dark” design elements, while other, more subtle manipulations remain unaddressed, is an ineffective strategy that often comes too late. It is important to understand that the way an interface looks is an integral part of how Design Power is enacted. It is time to begin discussions about regulating design.

If you want to read everything in more detail and with further thoughts, discussions and (most importantly) sources, here’s the link to the paper in the ACM database again:


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s