Dark patterns: Manipulative web design and its effects on user behavior

When scrolling on Instagram, clicking on a cookie banner or booking a hotel – dark patterns can be found in all of these places. These are digital design elements that are intended to influence user behaviour in the interests of companies. Dark patterns often appear on websites with the aim of obtaining money, data or the attention of users. The design of the individual patterns varies greatly: from the different colour design of the selection options in cookie banners to the supposed scarcity (e.g. when booking a hotel “Only four left!”) to influencing through language (“Are you sure you want to miss out on our unique offer just for you?”) or designs that make us scroll further and further in social networks (so-called doomscrolling). What all dark patterns have in common is the digital business model: the seemingly free or low-cost use of services such as online newspapers, YouTube or shopping websites is not paid for directly with money, but with data or attention. The distortion of financial offers (such as the artificial scarcity: “Only four left!” when booking a hotel) is also a business practice of dark patterns.

Dark patterns start at different points in the decision-making process. Basically, every decision initially consists of two options, for example accepting all cookies or rejecting all cookies. Each of these decision options has an individual value that people use to determine which option they prefer. This preference can depend on their own motivation (e.g. to protect their own data), the information available or also on existing time pressure. The value of the individual decision options can be changed: Options can be upgraded or downgraded. This is precisely where dark patterns come in, although they influence the decision options in different ways: Some dark patterns increase the value of an option, for example through an eye-catching visual design or by creating the illusion that many other people would also find this option great. Other patterns, on the other hand, reduce the value of options, for example by using emotionally coloured language to discourage users from cancelling a subscription. This pushes users in one direction when making a decision. Dark patterns can be understood as a sub-form of nudging. Nudging generally describes the modification of a decision-making situation to reinforce or initiate behaviour (e.g. labelling food with the Nutri-Score to promote healthy purchasing decisions) and can therefore also correspond to the consumer’s actual intention (e.g. to eat healthily). On the other hand, design elements are labelled “dark” if they are primarily geared towards the intention of companies. However, as people’s intentions are difficult to predict and can vary greatly from person to person, it is not always possible to clearly distinguish between nudging in the interests of users and dark patterns.

It can therefore be helpful to consider the context of digital decisions. When visiting a website, many decisions have to be made in quick succession: Where do I look? Which information do I trust? Which product do I want to buy? Not all possible options can be carefully weighed up in this multitude of decisions. In many cases, decisions are made intuitively. Certain cognitive simplifications (biases) come into play here, for example the orientation towards predetermined values (default bias). These simplifications therefore represent a kind of interpretation aid and enable quick decisions to be made. This is often helpful and protects against cognitive overload, but it can also lead to misinterpretations. Dark patterns exploit this mechanism and aim to make a quick decision based on misconceptions. For example, the message that there are only four products left creates the impression that an offer is very scarce and encourages a quick reaction that may not correspond to the user’s actual intention.

As a result, dark patterns make it difficult to make a self-determined decision. In addition to the potential costs incurred in the form of money, data or attention, studies also show that this can lead to negative emotions or a feeling of loss of control. If users notice the dark pattern and want to avoid the attempt to influence them (e.g. rejecting all cookies despite the colour design of a cookie banner), this can be accompanied by increased cognitive effort and loss of time and sometimes even lead to an increased heart rate.

Comparability with analogue phenomena

Business practices for influencing the behaviour of users can be found on a massive scale in the analogue world. Products in the supermarket are sorted in such a way that attention is drawn first to the more expensive offers, emotionally charged language is used in sales pitches, the route through shopping centres is often designed to lead past as many products as possible and advertising is not always directly identifiable as such. The change in decision-making behaviour through information filtering, attention control, emotional influence or increasing the density of stimuli and thus the complexity of a decision-making situation therefore exists both analogue and digitally.

However, dark patterns utilise these mechanisms in a different quality. Firstly, the stimulus density in digital decision-making situations is often higher than in comparable analogue settings anyway. Accordingly, the decision-making situation is more complex. Digital decisions are made more intuitively and quickly, which also makes them more susceptible to being influenced by dark patterns. In addition, the presentation of offers or information in the digital space can be more efficient and targeted towards users. In particular, websites with many users (such as large social media websites or platforms like Google or Amazon) have access to immense amounts of data that can be used to predict the preferences and vulnerabilities of certain consumer groups. Companies can use this knowledge to design websites in such a way that they are tailored to individual users and content is displayed in a personalised way. The high degree of possible personalisation in the digital space means that individual preferences can be taken into account and products can therefore be placed particularly efficiently. This goes far beyond tailoring certain offers to a target group in the analogue space. People have now become accustomed to many phenomena of analogue behavioural influence, have recognised the deceptive or manipulative potential and have developed strategies for dealing with them. On a digital level, this process is still ongoing. Studies show that there are already some habituation effects and people no longer fall for dark patterns and instead, if possible, even avoid the website in question. It remains to be seen to what extent people will become accustomed to digital influence. However, considering the complexity of digital decision-making situations and the data-driven personalisation of content, it can be assumed that dark patterns exploit other weak points in decision-making than would be possible in the analogue space.

Social relevance

Who should have the authority to decide what we read, buy and what information we use? The question of digital freedom of choice, of self-determined decisions, is also being negotiated in dark patterns. At a micro level, the potential influence of dark patterns may seem less relevant: staying on Instagram a little longer or accepting all cookies rarely has a noticeable impact on our lives. However, if we look at the macro level, these small decisions determine major processes. Especially in times when the internet is monopolised by large platforms, they can control the behaviour of many people through the design of their websites. If many individual users release some personal data every day or spend even just five minutes longer on a website every day through the design of a website, this results in a considerable amount of “new goods” in total. The question of who has control over our resources in the form of money, data or attention is therefore becoming increasingly relevant.

---