Trust & Safety policies are the guardrails that keep users safe on a platform. While the method of researching, writing, and operationalizing policies may differ by platform, policies convey the values and expected behaviors of a platform’s community. A well-executed policy will help a platform articulate what is, and what is not considered harmful behavior on the platform. Looking at data, appeals, user reports, and user retention will help inform platforms on how well the policies are written.
Agnes Evrard, Director, Trust & Safety, Brainly
Alex Sandeford, Product Policy Lead - Trust and Safety, Coinbase
Daryl Sando, Sr. Trust & Safety Policy Manager, Tinder
This past week, our #TSCollective Community Manager, Matt Soeth, sat down with Agnes Evrard, Director, Trust & Safety, Brainly; Alex Sandeford, Product Policy Lead - Trust and Safety, Coinbase; Daryl Sando, Sr. Trust & Safety Policy Manager, Tinder.
Main Take-Aways
- Think about your product and brand; what it accomplishes, what the product stands for, and is trying to accomplish.
Today, there are not a lot of new policies to write. We have frameworks to get us started whether they come from the early days of Yahoo or Google, there is a starting point. What every policy writer needs to take into account is to start with the principles and values of the product. Agnes encouraged writers to think about the age of the users, what type of environment you hope to create. Alex built on this highlighting policies will differ by platform depending on the values and the community norms. Policy is the action that brings your values to life and is a reflection of the company. - Welcome feedback from stakeholders in your company on the policy you have drafted.
An important step to implementing policy is to engage internal stakeholders like engineering, marketing, leadership, and your moderation team. There is a process in place when it comes to moderation workflows. Knowing that process, who is involved in the decision-making, and what information those stakeholders need to know in order to take action is a crucial step in getting buy-in. - Write enforceable policy - a policy that is actionable, clear, and agile.
Daryl Sando spoke about her time as a policy writer in government vs being a policy writer in tech. For tech, you have to write policy in a way that it’s meant to be used. Policy needs to be enforced by agents, so making sure they are able to enforce the policy and that it’s implemented consistently is a crucial step in writing the policy. Other skills that policy writers need in order to work in this field: data analytics, education, QA specialist and so much more. As a policy writer, you are constantly looking at data to see if the policy, as written, is effective in reducing the toxic behavior it addresses and supporting a resilient community. - Work with regional experts to have a policy that can be used on a global scale.
Another important skill is to have a global mindset when writing policy. Working with regional experts, and writing policy in a way that takes into account regional points of view will enable your policy to be more effective. Localization of policy takes input from a diverse team with knowledge of the regulation, norms, and values of the local culture. As much as you are able, engage with these groups to ensure that policy is localized and actioned correctly and consistently. - Work with your data analytics team, product team or an outside analytics vendor to understand the impact of world events on user behavior.
One of the biggest challenges that policy writers face is writing new policies when none exists. This is hard to measure, because if it doesn’t exist, then it is not being measured or categorized by your moderators as they don’t know to look for it. Strategies to solve these difficult issues is to look at themes that are emerging on your platform. These themes may even be surfacing in media sources. Working with data analytics or product and process and doing keyword research for these themes in the platform for types of behavior and prevalence. Now, you can identify if a policy needs to be written. In addition, you can do searches on similar platforms to see if these themes and behaviors are prevalent in other spaces as well as determine if this issue is just on your platform or if it’s industry-wide. This will inform if you need to write a policy for behavior on your platform, or off your platform, that needs addressing to keep users safe. - Have a rollout plan before implementing a new policy.
One big key takeaway from this panel conversation, have a launch plan. For any policy, know how you want to roll it out. Start with internal education, work with your moderation team, and scope out a policy as far as you can through case banking. Case banking is a collection of evidence, posts, or content that shows a missing gap in your current policy. This is a great way to show your stakeholders the need to change and adjust policy. As a policy writer, if you have done your research, you will know the risk profile of a specific behavior; be able to communicate the impact on the brand, and impact on your members, as well as identify what legal responsibilities you have as a platform. - Educate your users on the platform's community guidelines.
Last, writing internal policies vs external community guidelines is a different skill set. Most users do not read the guidelines until something has happened. Embed education into your product, and make it part of the user experience. Communicate clearly, and broadly, what your policy enforcement will do. While internal policies may be constantly shifting and changing, external policies are much more stable, broad, and change less often. Your goal is to educate the user whenever possible and be transparent as possible with your decision-making and enforcement. If you take a stand on policy one day, do not change it a few days later. Being consistent is a major key to success.
When it comes to tech policy writing, the biggest thing to remember is that policy is action. You are operationalizing the ethos of your company in order to create a great user experience and build a resilient user community. It’s an ongoing process that never slows down and is informed by data, and human behavior. As such, policy is the foundation for everything we are trying to do as a platform and represents everything we hope to do for our users.
Additional Resources
-
Watch the master class on-demand.
-
Sign up for #TSCollective, a community of trust and safety professionals sharing best practices and support for this heroic work.
- White Paper | Developing and Operationalizing Policy
- Spectrum Labs | Community Health Assessment
-
Spectrum Labs uses Contextual AI and Trust & Safety expertise to provide you with a report and analytics of the behaviors on your platform. Contact our team for a community health assessment.
-
- Spectrum Labs | Analytics & Reporting
-
Spectrum Lab's configurable AI adapts to changing behavior and continues to improve over time. It gives Trust & Safety teams insight into new behavior that may arise from current events. Contact our team to learn more.
-
