Towards a greater emphasis on accessibility within the trust & safety industry

Christine Lehane, PhD
UX Collective
Published in
7 min readJul 10, 2022

--

Image shows a keyboard where one key is green and has a lock symbol on it. The shift key to the right of it is blue. All other keys are light grey.

TL;DR | Research highlights that people with disabilities report greater exposure to harm online but that products designed to prevent or remediate online harm suffer from accessibility issues. In this article, I outline the T&S Accessibility Action Plan which is a straightforward process that can be used to improve the accessibility of front-end products.

As I’ve been gathering resources for my Social Science to Trust and Safety Series, a publication titled “Accessibility for Trust and Safety Flows” caught my attention. It caught my attention because a) my academic research primarily focuses on adjustment to sensory loss (i.e., vision and/or hearing loss), but also because b) as the article shows, online trust and safety products are commonly inaccessible for people with disabilities. This is a problem. Accessibility is a vehicle for inclusion and inclusion is one of the core values of online trust and safety systems. Without inclusive design, products developed by trust and safety teams will primarily protect people in majority groups, neglecting those who may have a greater need for support online such as people with disabilities (Alhaboby et al. 2019).

Access to online safety tools is becoming increasingly important. Digital reliance is increasing both in the general population and among people with disabilities (Martiniello et al. 2022). Especially in the past two years, more people around the world have become reliant on online services but online services are not always built with everyone in mind. This is also true for the trust and safety industry.

As people broadened their use of online services and spent more time online during the pandemic, reports of online scams, harassment, hate, and other privacy and safety concerns greatly increased (Bitaab et al. 2020; FTC, 2022; World Bank, 2021). While it is hard to know whether this increase in online harm disproportionately impacted individuals with disabilities, we do know that studies conducted prior to the pandemic generally found a higher self-reported prevalence of harmful online experiences among people with disabilities compared to those without (Alhaboby et al. 2019).

Reports of greater exposure to online harm among people with disabilities clearly suggest that more tailored harm prevention work needs to be done within the trust and safety industry. However, what I find personally more concerning is the research showing that products designed to help users remedy or prevent harmful online experiences suffer from accessibility issues (Thiel & Bradshaw, 2022; Lobo et al. 2017). Why is that? As most people working in this industry will know, it is an impossible task to build a system that will capture and prevent every instance and form of online harm. Many harms are subjective, some are interpretable only to those with full context, and perpetrators continuously evolve and obfuscate their actions. For this reason, accessible, usable, and effective safety tools are crucial to support self-remediation where proactive trust and safety systems are insufficient.

With this article, I suggest a process by which trust and safety teams can begin to identify and address accessibility issues within their products — the T&S Accessibility Action Plan. While I acknowledge that biases in backend system design can disproportionately impact different user groups, in this article, I am focusing exclusively on front-end products.

The T&S Accessibility Action Plan

Step 1: Identify Axes of Inaccessibility

Firstly, while I approach accessibility discussions through the lens of visual and hearing impairments, that is far from the full picture. Accessibility is about making sure anyone can use and benefit from the same product, irrespective of their physical, sensory, cognitive, socio-cultural, or economic circumstances.

Vision and auditory functioning are two of what I call “axes of inaccessibility”. These are factors that, depending on your level of impairment, can determine how accessible or inaccessible a product is for you. Other examples of axes of inaccessibility include gender, age, digital literacy, race, geographic location, residence status, etc.

A good first step when creating an accessibility action plan is to conduct desk research, brainstorm, and familiarize yourself with the factors that can make a product inaccessible. This will not only support your current product development process but will provide a foundation of knowledge for any future product ideation.

Sticky notes in yellow, green, and orange with one axis of inaccessibility written on each sticky note. Axes mentioned include Gender, Literacy Level, Digital Literacy Level, Status of Residence, Visual Function, Age, Race, Auditory Function, Dexterity.

Step 2: Audit the Product Experience

Now that you’ve created your list of axes of inaccessibility, the next step is to identify which ones are the most relevant for your product. Using a realistic prototype or live product, conduct a walk-through analysis and document what capabilities and resources a user would need to be able to proceed through each step of your product experience. For example, if you are designing a QR code-based authentication system, some relevant axes of inaccessibility would include visual functioning, digital literacy, and device type. If you are designing a government ID-based authentication system, some relevant axes of inaccessibility would include age, gender, geographic location, dexterity, visual functioning, and device type.

You may also use a color-coding system (e.g., traffic lights) so that at a quick glance, you can identify the most accessible or inaccessible steps in your user journey.

Image shows a user journey map from steps one to five. There is written guidance to document the capabilities and resources needed for each step of the user journey.

Step 3: Optimize the Design

For an optimal user experience, you should aim for low product dependency on your relevant axes of inaccessibility. If you have two design ideas where one is highly subject to three axes of inaccessibility while another is only highly subject to one, all else equal, the second design is likely preferable. If you only have one design idea and it is highly dependent on many axes of inaccessibility, you should consider drafting design alternatives that are more broadly accessible.

Additionally, if you are aware of the prevalence of each of your core axes of inaccessibility within your user population, you may choose to optimize your product to improve accessibility for the maximum number of users.

This image shows a sample grid that can be used to compare designs on different axes of inaccessibility. In this grid, the product name and image is placed on the left while the axes are rated from low dependency to high dependency on the right.

Step 4: Conduct User Interviews

Steps 1–3 are great prep work but to know whether or not your product is accessible, you need to see how people use it. Using the axes of inaccessibility that you identified as most relevant for your product, you should invite people with those specific accessibility needs to participate in user interviews. Using a realistic prototype or live product, you can conduct user testing sessions to better understand the accessibility of your product, including pain points and optimal design patterns. These sessions can also help you to understand why parts of your design may be inaccessible and help you to avoid problematic design patterns in future product development processes.

It is important to note that when conducting research with people with accessibility needs, everything from the recruitment process to the one-on-one interview setting will need to be accessible, otherwise, you may bias your sample and not get a true picture of the adaptations needed for your product. For an in-depth guide on accessibility-focused user interviews, see this report from the Nielsen Norman Group.

Alternative / Supplementary Steps

Realistically, for a variety of reasons, conducting accessibility sessions with users is not always possible. Your team may not have the budget, a designated researcher, time, etc. Worry not, there are some lightweight options that can be used to catch at least some of the accessibility issues with your product.

  1. Similar to the common UX research technique, Heuristic Evaluation, you can conduct a walk-through analysis of each part of your product experience but instead of documenting its general usability issues, you can use apps to test how accessible it is. If you are interested in understanding accessibility needs for users with visual impairments, there are apps such as the NEI VR: See What I See app that can show you how a person’s visual field is affected by different medical conditions.
  2. Another alternative is to partner with accessibility experts or expert organizations who may advise your team or schedule a product audit for you. There are several such organizations including the Hearing Loss Association of America, the American Foundation for the Blind, and Deafblind International.

Closing Remarks

Accessibility is generally important for digital products but it is especially important for the trust and safety industry as evidence suggests that people with disabilities and other minority groups are disproportionately affected by online harms such as harassment, hate, and scams. Proactive harm detection systems are not sufficient to mitigate all harmful online experiences, thus, it is imperative that tools built to prevent harm or remediate it when it has occurred are accessible for all users. Many companies already have accessibility testing as standard practice, especially for trust and safety products. In such cases, publishing this work should be encouraged to serve as inspiration for the field. Finally, accessibility testing does not have to be very complex, even in a resource-constrained setting, it is still possible to conduct lightweight accessibility tests or to consult with expert organizations and improve your product experience.

--

--