“For Schibsted’s media business, we have concluded that trust is the most important asset in the increasingly automated and fragmented information landscape. Understanding the dynamics of trust in these times of change is not just a matter of curiosity – it’s a necessity for our business model and the social function we seek to uphold,” says Siv Juvik Tveitnes, EVP News Media and future CEO of the new media company Schibsted Media.
She points out that trust is often discussed as a core value and ideal for editorial media, but perceptions about what the term represents tend to be broad and varied. Therefore, a unique study was undertaken during the fall of 2023.
“We wanted to gain a valid, nuanced, and operationalizable definition of trust and its drivers. We aimed to ensure that trust isn’t something that we in the media industry are working with from the inside-out, but rather a topic where we put users’ attitudes first”, says Fredric Karén, responsible for editorial transformation and collaboration in Schibsted News Media, and owner of the trust study.
After gathering various types of insights – including academic research, internal and external expert interviews, and focus groups – a large systematic survey study with 3,000 media users aged 16-74 in Sweden and Norway was conducted, in partnership with NoA Consulting. Utilising Norstat panels, a participant mix representative of the national populations was sampled.
Identifying key drivers
Many different media attributes were tested in the study, and after statistical analysis they were grouped into categories. Among these, the following top four drivers were identified:
- Credibility of process
How credible the process and people behind the content is - Credibility of content
How credible the content is in itself - Personal relevance
The perceived relevance and usefulness of the content, for the individual user - Selectivity
What facts, events, and topics that are covered (or not)
“While some may connect the most important driver –‘credibility of process’ – to human editorial processes, it’s important to also consider how our use of emerging technologies can contribute to enhancing media trust”, says Agnes Stenbom, project lead for the trust driver study.
Key difference between Sweden and Norway
The drivers were highly similar across the sub-groups analysed in the study. Norwegian and Swedish users were 94% similar in terms of which attributes make them trust a media brand. A similar high alignment (89-100%) was seen when comparing users based on e.g. age, gender or educational level.
But interestingly, there was a significant difference in the importance of the specific attribute ‘Accountable editor’ in Sweden versus Norway. Among the Norwegian respondents, this was the number one driver; the single most important attribute driving trust. Among the Swedes, the same attribute was ranked as number 19.
“We believe part of this difference can be attributed to the varying terminology used to describe editorial media in Norway and Sweden, respectively. In Norway, the category is often called “redaktørstyrte medier”, directly translated to “editor-steered” or “editor-led” media. While similar words exist in Swedish, like “redaktionella medier”, they are seldom used outside of the industry. Instead, Swedes talk about “journalism” or “the media” – which doesn’t necessarily bring to mind the roles and responsibilities of the editorial process”, Stenbom explains.
While language may play a role in the difference between Swedish and Norwegian perceptions’, there are other potential explanations to consider.
“The accountable editor is more of a public figure in Norway, not least with Pressens Faglige Utvalg (PFU) hosting public questionings”, says former editor Fredric Karén.
Trust and willingness to pay
The study also explored how trust impacts the extent of usage and the willingness to pay for content from editor-led media.
“Interestingly, we found clear signals of trust and usage not being driven by the same attributes. Or phrased differently: what drives trust does not necessarily drive usage. A key example can be found in how the attribute ‘Accountable editor’, which is a key driver of trust, seems to have very limited impact on usage. In contrast, ‘User friendly’ is an attribute with great impact on usage but is seemingly unimportant for trust”, says Stenbom.
She stresses that this doesn’t imply Schibsted’s media houses should cease creating user-friendly products. However, Stenbom emphasises the importance for media companies to prioritise and focus by understanding which attributes are likely to drive particular behaviours or attitudes.
“The other attitude we tested for was willingness to pay. Here, there are much clearer synergies with the attributes that drive trust. What is important for trust seems to be equally important for willingness to pay. This is an important finding for any media looking for sustainable ways of financing their independent journalism. Investing in trust can be key to unlocking user revenue”, says Stenbom.
Deep dive on personal relevance
When comparing drivers of trust, usage and willingness to pay, ‘Personal relevance’ emerged as a critical factor driving all three. Unpacking this driver, the study shows that the attribute of ‘Aligns with my worldview’ is key, and interestingly, more important than providing diverse perspectives. This was also a relatively more important driver of trust for consumers aged 50 and older.
“This can of course be interpreted in different ways, and we surely need to do more research to understand how people understood the question. But it is not hard to see signals of this tendency playing out in other media markets. Looking west, we can for example see how newsrooms with a clear political stance tend to gain great trust among their own users – and quite the opposite among those who hold other political beliefs”, says Stenbom.
She highlights that Schibsted has been betting on personal relevance for a long time, not least with the work within algorithmic personalisation of editorial content.
“As communicated through a recent article on Schibsted.com, our system ensures a major story always trumps personal interest. Even if it might provide short-term wins in user trust, we believe that always aligning with or confirming users’ opinions or perspectives is an unattractive path forward for editorial media. In an era of (possible) hyper personalisation, we can play a key role in safeguarding that there is some sense of a shared story about our time,” says Stenbom.
The road ahead
Siv Juvik Tveitnes explains why it makes sense to share the findings from the trust study with the entire industry, rather than keeping them internal.
“It is our conviction that the media industry benefits from open discussion about trust. We all thrive when there is trust in editorial media, and similarly, we can be collectively punished when trust is lost in one brand. We are therefore eager to discuss these findings with our industry peers,” says Siv Juvik Tveitnes.
Agnes Stenbom has recently been assigned to lead initiatives centered around the identified trust drivers, a role she will maintain as Schibsted News Media transitions to becoming an independent company named Schibsted Media during the second quarter. With a background working with artificial intelligence and inclusion in and beyond Schibsted, she will work closely with newsroom representatives to facilitate an exchange of learnings and identify initiatives for collaboration.
“We expect there to be numerous new initiatives across our group as we explore ways to evolve credibility of process and content, personal relevance, and selectivity. Trustworthy media has never been more important, and we are eager to continue our work with safeguarding it,” says Stenbom.
| About: How the key drivers were identified |
|
The purpose of the study was to understand what drives people’s trust in media at a category level (”editorial media”). Questions to users were, however, asked at brand level to evoke more nuanced and informed answers, compared to questions about “media” or “news media” as a whole. The brand-level answers were then analysed at an aggregated category level. Direct questions about what users think or feel about trusting a media brand were avoided to prevent rationalised answers influenced by social-desirability bias. Instead, statistical analysis was employed to identify ”true” drivers by correlating and finding links between different attitudes expressed. In the initial part of the study, respondents rated on a 1-7 point scale the extent to which they trust information from a specific media source. Later, they were presented with various statements about media attributes (e.g., ”Here, information is fact-based and true”) and asked to assess their truthfulness for different media sources. These responses were then analysed to correlate these questions and identify key drivers of trust. |
