Researching people’s experience of harmful video content

Sector: Digital

Partner: Oliver Wyman

Diagnostic survey

Researching people’s experience of harmful video content

We developed a behavioural user survey that helped us gain insight into the potential harmful video content young people are exposed to on social media, and if and how they would go about reporting this.

The Challenge

In recent years, concerns have grown about people - especially young people - encountering harmful content while using online video sharing platforms such as YouTube, Facebook, TikTok, Instagram and Twitch. The Department for Culture, Media and Sport wanted to get a better understanding of the potential harms encountered by young people through the development of more rigorous methods for understanding their experiences. We took on this challenge as part of a wider programme of work with Oliver Wyman, the global management consulting firm.

The Approach

In order to get a better understanding of the potential harmful content encountered by young people, we developed a ‘behavioural user survey’. The core insight at the heart of this method is not to restrict responses to a specific list of harmful content types, but rather give participants the opportunity to describe whether or not they had encountered content that they believed was ‘inappropriate, distressing or intentionally misleading’. The aim was to avoid the problem with more traditional surveying methods, which tend to constrain participants’ by asking them to select a response from a pre-defined list. This also had the benefit of ensuring that the context in which a respondent encountered a particular form of content could be taken into consideration. Footage of a medical procedure, for example, might not be seen as being harmful in the context of an online healthcare documentary; but might be perceived as being harmful if seen unexpectedly and without consent in a social media feed.

To run the behavioural user experience survey we recruited 2,252 UK users of video sharing platforms. We wanted to make sure we heard from young people too, so within our sample were 355 young people aged 18 or under. Respondents were asked to recall and describe (in an open ended question) one specific instance of encountering video content that they considered to be inappropriate, distressing or intentionally misleading. They were also asked how often they encountered content of this kind, what they did when they encountered it, and what happened if and when they reported it to the platform operators. In other words, we were interested in the extent to which action was taken (by them and the platform owners) if harmful content was encountered.

The Result

Almost half (45%) of the respondents to the behavioural user experience survey reported having encountered inappropriate, distressing or deliberately misleading content. The most commonly reported forms of content were violent or disturbing videos (19%); fake news and disinformation (18.7%); inappropriate sexual content (8.2%); and animal abuse (7.5%). 1% reported content containing sexual abuse and 0.7% reported content promoting self harm.

One of the key findings from the survey was how little action is taken upon encountering harmful content. Most people who encountered harmful content did not take any action in response (822, ~55% - see chart below). The most cited reasons were a belief that taking action would not make any difference (49%), or the opinion that the content they encountered was not harmful enough to cause them to take action (29%).

Figure 1: A diagram showing the distribution of users’ reactions upon encountering harmful content

(C) CogCo

Of the participants who had reported the harmful content to the hosting platform, many (42%) did not know what subsequently happened. This suggests that video sharing platforms are not always closing the loop by informing those who make the reports of the actions they take. We know that feedback about past actions can help encourage future engagement, so this lack of information from platforms might have the effect of deterring future reporting. We also found that only a small minority of people believed that other users actually report harmful content. This is important, as we know from other behavioural studies that people are more likely to do something when they perceive it as the ‘normal’ course of action.

This survey has helped to contribute to the wider work being undertaken by government and industry around online harms. And we believe that the methodological approach of running surveys in this more sophisticated way will help to plug gaps in knowledge in many other areas of research.

“It was great working with CogCo in this important area. The behavioural survey at the heart of this project has helped to contribute to the wider work being undertaken by government and industry around online harms.”

-Lisa Quest, Partner, Head of UK and Ireland, Co-Head of the Public Sector and Policy Practice Europe (Oliver Wyman)

Read recommended case studies

Understanding which elements of a phone package customers value most

Mobile phone operators have a wide variety of services and products that they can offer to their customers. They can provide more data; better coverage; shorter contract terms; and even free gifts, like a 6 month subscription to Netflix. But each of these elements comes at a cost, so it is useful to know exactly what value customers place on the different component parts of a contract. This might not be a problem, if customers were able to articulate clearly their preferences. But it turns out that we find it very difficult to do so accurately: it is challenging, for example, for any of us to come up with a figure when asked how much you might be willing to pay for a shorter contract.

Understand behaviours

How digital design changes customer decisions

Online shopping offers customers a smooth, simple way of making a purchase. But Citizens Advice, the charity that offers confidential financial advice to millions of people every year, is concerned that some firms are designing online platforms in ways that act against the interests of customers. Citizens Advice therefore commissioned CogCo, working with Paul Adams, to devise an experiment that tested what the effect of different aspects of digital design had upon the choices and comprehension of customers. They were especially interested in the effect of different digital designs upon consumers’ decisions to choose to pay using ‘Buy Now, Pay Later’ products. These products enable customers to delay payments into the future and now account for 6% of all online purchases in the UK.

Helping More People to Save for a Rainy Day

Around a third of all people in the UK have less than £100 in savings. This means that tens of millions of people across the country have next to no financial buffer, should they encounter any form of financial challenge. A broken fridge, or an unexpected parking fine, for example, can lead to someone needing to borrow money that they subsequently find it difficult to repay. The team at Wagestream, a financial wellbeing app whose goal is to make money management easier and more accessible for people, recognised the severity of this problem amongst the 500,000+ users of their platform. They asked CogCo to work with them in helping them to build a new product that could help their users build up their financial resilience.