Have you ever clicked ‘Yes’ when asked 'Have you read our terms and conditions?' even when it wasn’t true?
We wanted to find out if behavioural techniques could be used to increase the number of users accessing online terms and conditions (T&Cs). And does accessing T&Cs actually make a difference to how people behave online?
To read or not to read…
The question 'Have you read our terms and conditions' induces a mild internal conflict in me. Part of me wants to be a responsible online citizen, fully informed about the type of services I’m using, aware of any rules for using that platform or find out how they plan to use my data.
Another part of me really can’t be bothered. Do I really have the time to scroll through pages of documents written in complicated language? Besides, what’s the worst that can happen? Surely the service isn’t allowed to do anything illegal. So, more often than not, I click ‘Yes’, hope for the best and become complicit in what's become known as the biggest lie on the internet [1].
It turns out, I’m not alone. Our survey research shows that between half and two-thirds of users reported signing up to online platforms without trying to access or read T&Cs. We also found that people tended to avoid T&Cs and platform rules because they thought they would take too long to read, they would find them overwhelming, or they were confident they were not going to do anything that violated the rules.
Our analysis shows that some of these concerns might not be unfounded. We found that terms of service for popular video-sharing platforms (VSPs) can take anywhere from eight minutes to over an hour to read, with most requiring advanced readings skills to understand.
However, social media terms and conditions, and specifically community guidelines, include important information about how to stay safe while using the platform. This means they act as an important starting point for protecting users from harms and creating safe and trusted online communities. However, in order to do any of these things, users need to engage with them.
Did you know?
To demonstrate how few people actually read their privacy policy, a company recently revealed that they had included a hidden clause for a free bottle of wine buried within it. It went unnoticed for months but was eventually spotted and claimed by an eagle-eyed viewer.
Sometimes it pays to read the fine print!
Our experiment
We used a simulated social media platform (called ‘WeConnect’ designed by the Behavioural Insights Team) to test whether behavioural techniques could increase the number of people accessing the community guidelines.
The interventions focused on how (i.e., changes to the user interface) and when (i.e., at sign-up or while scrolling) the link to the community guidelines was presented.
We tested:
- a control. A link to community guidelines was presented to participants while they were signing up to the platform. This was intended to represent common practices by platforms;
- a reframing message. The link to the community guidelines was presented while users were signing up to the platform alongside a motivational message to highlight their importance. Our survey suggested that users do not perceive T&Cs or platform rules as valuable to them or their online experience. We used a reframing message to try target these concerns and increase motivation to access the community guidelines;
- a relabelled title. The name of the community guidelines document was changed to be more user-friendly (‘Dos and Don’ts’) and presented while signing up to the platform. Our survey suggested that only one in five users listed community guidelines as a place they would go if they were unsure about posting something on a platform. We wanted to test whether giving them a more user-friendly name, that clearly communicated the purpose and contents, could increase the number of users accessing; and
- prompts. Whilst participants were scrolling the content feed, they saw an active prompt that either included the reframing message or the relabelled title. To our knowledge, there is limited encouragement from social media platforms to read their T&Cs or the platform rules. We wanted to test whether providing a prompt to users at a more salient time point (such as when they had started using the platform and been exposed to content) could encourage more users to access community guidelines.
Screenshots of the different ways the community guidelines were presented to users on WeConnect
We were also interested in whether any of these interventions impacted how people subsequently behaved and interacted with content on the simulated platform. So, we measured the number of participants that reported or reposted ‘violative’ content. (i.e., content that is explicitly prohibited in the community guidelines). This included content involving violence, hate speech or misinformation.
WeConnect’s Community Guidelines
Prompting can help, but isn’t a magic bullet
We found that, out of our interventions, prompting was the only effective way to increase engagement with the community guidelines. Clicking to access the community guidelines rose from 3% in the control group to between 7% and 9% in the prompting intervention groups.
The percentage of participants who clicked to access the Community Guidelines
Note: ** statistically significant at the 1% level (p<0.01) in comparison to the Control arm.
Encouraging users to engage with T&Cs documents, at relevant points in time, outside of the sign-up stage, seems to lead to more engagement. Prompting could therefore address important opportunity barriers in accessing community guidelines (such as the community guidelines not being made salient to users). However, prompting is no magic bullet. Over 90% of users who received a prompt still chose not to access the community guidelines.
Why is it so difficult to get users to engage? It is possible that users’ motivation to access and read the documents is just very low for the reasons describe above (e.g., perceptions they will take a long time to read or are difficult to understand) and this is difficult to overcome with the types of techniques we tested. It’s possible that our reframing message was not powerful enough to shift motivation.
Our findings suggest that it will be important for platforms to consider other methods of encouraging compliance with T&Cs, beyond just increasing engagement. For example, one option might be for service providers to provide reminders of key information at the point of potential rule violation (such as uploading or reposting content).
Accessing community guidelines did not appear to change other online behaviours
We didn’t observe any significant change in the users reporting of violative content across any of the intervention groups. Between 19% and 23% of users reported at least one violative post with no statistically significant differences observed across the five intervention groups.
It could be that accessing rules and understanding what is and is not allowed on a platform is not the key determinant of how people behave online. For example, social norms and belief in effectiveness of reporting systems may have a stronger influence on these behaviours.
The percentage of participants who reported at least one violative post
Similarly, reposting of violative content ranged between 5% and 6% across all the intervention groups with no statistically significant difference observed.
It is worth noting that reporting and reposting were secondary outcome measures in our experiment so we cannot attribute causality between interventions and observed outcomes.
It’s worth recognising the limitations of our study - participants were aware they were part of a study, therefore, they might not have behaved as they would in real life. We discuss further limitations to the methodology in the discussion paper.
So…to read or not to read?
That is the question - but unfortunately, I still don’t have the answer. If you do choose to read T&Cs, you will learn important information about how to use services appropriately, keep yourself and others safe online, and who knows, maybe win some free wine!
However, our research suggests that it’s hard to get people to engage with T&Cs online, and even if you can, it still might not lead to the expected or desired changes in behaviour. Social media platforms cannot rely on having a written record of their T&Cs or community guidelines, which users can access, as the only tool to encourage desirable behaviours online. A variety of measures are likely to be needed to encourage users to comply with community guidelines, behave accordingly and create safe online experiences for everyone.
For further details on the research discussed in this article, read our discussion paper and technical report.
Zach Mills wrote this piece, based on research by Zach Mills, Ceren Cibik and Jonathan Porter with the Behavioural Insights Team.
Disclaimer: The analyses, opinions and findings in this article should not be interpreted as an official position of Ofcom. Ofcom’s Behavioural Insight blogs are written as points of interest and are the personal views of the author(s). They are not intended to be an official statement of Ofcom's policy or thinking.