Apply both a DEI framework, and an ethical framework, to the design of products and services

You don’t have to manage a product, to think about product inclusion. If you are not sure where to begin, first examine the institutions in which your product or service exists, identify the systemic barriers that exist within those institutions, and then proactively address them.

For example, many tech companies—from lenders to real estate to insurance—rely on credit scores. But what many teams don’t consider is the well-documented effect of generations of racial bias on credit scores, and how that systemic bias might be affecting users.¹ Therefore, any product that relies on credit scores has an opportunity to apply an inclusion lense to their work.²

Every product and service connects to larger institutional barriers. For example, if your company focuses on education, consider how inequitable funding of public schools impacts your customer and product. Or perhaps your product inadvertently treats customers who have been through the justice system differently. When product managers actively consider barriers in the systems where their products operate, they can develop a DEI lens and apply their understanding to their businesses.

1. Inclusion and accessibility product work requires more of a time commitment and is demonstrably greater than other types of product work. If you are strapped for the necessary time or resources, consider postponing your efforts because, when executed poorly, this type of work has more potential to cause harm than other product work.
2. Product team members should proactively educate themselves on general accessibility barriers, inclusivity challenges, and current literature/best practices. It is important to stay up to date with and use inclusive language and preferred terms.
3. Inclusive product strategy should not center on requiring users to identify themselves as needing particular accommodations within the product or specific features, as this risks creating siloed experiences. Instead, focus on embedding product changes that increase inclusion/accessibility overall and providing easy paths to access for all groups.

Ashlee Edwards, Ph.D., Product Inclusion Research Strategist (currently at Snap Inc, formerly Netflix, Instagram)

WCAG’s approach to creating accessible web content is based on four foundational principles: perceivable, operable, understandable, and robust (POUR). The guidelines provide detailed guidance on each of these, but W3C provides the following summary on its website:¹

PERCEIVABLE

• Provide text alternatives for non-text content.
• Provide captions and other alternatives for multimedia.
• Create content that can be presented in different ways, including by assistive technologies, without losing meaning.
• Make it easier for users to see and hear content.

OPERABLE

• Make all functionality available from a keyboard.
• Give users enough time to read and use content.
• Do not use content that causes seizures or physical reactions.
• Help users navigate and find content.
• Make it easier to use inputs other than a keyboard.

UNDERSTANDABLE

• Make text readable and understandable.
• Make content appear and operate in predictable ways.
• Help users avoid and correct mistakes.

ROBUST

• Maximize compatibility with current and future user tools.

GLAAD released its Social Media Safety Index in May 2021, outlining general recommendations relevant to social media platforms around LGBTQ+ self-expression, privacy and outing, and mitigating anti-LGBTQ+ hate.¹

KEY TAKEAWAYS

• Stop allowing algorithms to fuel extremism, disinformation, and hate.
• Similarly, confront the problem of bias in AI, which disproportionately impacts LGBTQ people and other marginalized communities. Consider using more human moderators who are appropriately trained to identify cultural biases.
• Make it easier for users to report problematic content, and be transparent in content moderation.
• Employ a dedicated LGBTQ policy lead, and take leadership in civil discourse.
• Respect data privacy, especially where LGBTQ people are vulnerable to serious harm and violence.
• Only select platforms currently take action on violent speech and misinformation, usually by monitoring trending topics for misinformation, restricting hashtags or shares, or posting labels on misinformation. When it comes to anti-LGBTQ misinformation, enforcement is arbitrary at best.

To go beyond general guidelines, the Center for Equity, Gender, and Leadership (EGAL) at the Haas School of Business at the University of California, Berkeley released a playbook that provides a comprehensive bias in AI map and notes specific actions that business leaders can take to effectively mitigate bias in AI technology. The playbook outlines “strategic plays” in three areas: teams, AI model, and corporate governance and leadership:

TEAMS

• Enable diverse and multidisciplinary teams to research, develop, operationalize, and manage algorithms and AI systems.
• Promote a culture that encourages employees to prioritize equity throughout the process of developing AI. Be transparent about shortcomings.

AI MODEL

• Practice responsible data set development, including by gathering inclusive data.
• Establish policies and practices that actively check and mitigate bias in algorithm development.

CORPORATE GOVERNANCE AND LEADERSHIP

• Establish corporate governance for responsible AI and end-to-end internal policies to mitigate bias.
• Engage corporate social responsibility (CSR) teams to advance responsible AI and larger systems change.
• Use your voice and influence to advance industry change and regulations for responsible AI.

For more information and how-to guidance on the strategic plays, as well as quick wins to take advantage of, explore the playbook from the Berkeley Haas School of Business.¹

Also see the Algorithmic Justice League’s Equitable and Accountable AI ²

The World Economic Forum, in partnership with BSR, released a framework for responsible innovation that combines ethics-based and human-rights approaches and applies them at every stage of the process, from initial design to customer experience. The framework breaks down the life cycle of disruptive technology into three main phases:

1. How they are designed and developed
2. How they are deployed and who they are sold to
3. How they are applied and used by an individual or entity other than the original creator

At every phase, companies should be intentional about incorporating ethics and human rights-based approaches, and be clear about who is responsible at each step.

ETHICS‑BASED APPROACHES

• A framework for decision‑making in situations where right and wrong, good and bad, are not clearly defined.
• Address issues of fairness and social justice where different schools of thought and ethical standards exist—when various choices can be made and different paths can be chosen.
• Different traditions, cultures, countries, and religions may choose different outcomes and priorities suited to specific needs and sensitivities.

HUMAN RIGHTS‑BASED APPROACHES

• A focus on the experiences of the most vulnerable, and a holistic recognition of what all members of society need in order to live with dignity and thrive.
• Based on internationally recognized laws and standards; a common standard of achievement for all people. Established rights that should always be protected and respected.
• A minimum threshold and baseline expectation for the responsible use of disruptive technology.
• An internationally endorsed framework for defining company responsibility that considers the critical role of governments.

Excerpt from WEF and BSR’s white paper on Responsible Use of Technology ¹

The Unstereotype Alliance created a simple framework to help develop and evaluate creative content for progressive portrayals of all people. The guidance is designed to eradicate harmful stereotypes and bias throughout the entire creative process. It features three Ps – Presence, Perspective and Personality – and includes questions that can serve as a checklist of considerations when creating and evaluating content.

Presence: Focuses on who is being featured in the ad
Many kinds of people are completely absent from communications l—and to see your own image entirely erased in the culture that surrounds you is deeply disempowering.

• Who is portrayed in the communication? Who is the central character?
• Does the communication feature a range of people that are representative of the cultural and ethnic mix in the market?
• Do the characters go beyond being a ‘mannequin’ for the product?
• Has the character been taken into consideration int the context of the copy, clothing, positioning, naming, etc.?

Perspective: Focuses on who is framing the narrative
Who seems to be directing the action? Who is the person whose perspective is represented?

• Do we see the personal experience and perspective of the character?
• Who is driving the narrative?
• Is there any objectification or sexualization?
• Is this work challenging outdated perspectives in society about people/ communities/ relationships/ norms?

Personality: Focuses on the depth and agency of the character
Characters should have 3-dimensional personalities that feel authentic – funny, caring, strong, thoughtful, and respected.

• Do the characters come across as empowered and in control of their lives?
• Do the characters have 3-dimensional personalities?
• Is beauty a dimension of personality rather than just about physical appearance and attraction?
• Is there a stereotypical interpretation of beauty?

In 2021, Amazon Studios introduced an Inclusion Policy and Inclusion Playbook to improve representation both on camera and behind the scenes. The Inclusion Policy lays out Amazon Studios’ expectations for its internal teams and creative partners. It aims to tell stories that amplify the voices of characters across the spectrum of diversity — including race, ethnicity, nationality, sexual orientation, age, religion, disability (including mental health), body size, gender, gender identity, and gender expression. It also aims to create an inclusive production environment. The policy covers developing stories and characters, hiring and production, reporting and documentation, and a commitment to accountability. Amazon Studios will track data to hold itself and its partners accountable.

The Inclusion Playbook is a practical guide for the company and its partners on how to make good on these commitments by minimizing and disrupting biases at every step of the production process. People who engage with the Playbook will learn to:

• Make inclusive decisions, while telling authentic stories and hiring the best people for the job.
• Know where to go for help, including tools to make inclusive decisions.
• Identify criteria for making informed and consistent creative choices.
• Foster curiosity and ask questions to disrupt the status quo about how things are done or expectations about what storytellers want/do not want.

As part of this work, Amazon suggests “10 Gut Questions” to ask in any storytelling project:

1. Are we celebrating people of different ages, body types, skin colors, orientations, ethnicities, identities, and disabilities?
2. Do we showcase customers in a variety of relationships and roles?
3. Is there a diverse character for a primary role (e.g., speaking role vs. background)?
4. Does the diverse character feel tokenized or “othered”?
5. Is the diverse character defying stereotypes and respecting life experiences?
6. Is the diverse character in a position of empowerment?
7. Have we highlighted different environments, from urban to rural?
8. Have we represented people from different socioeconomic backgrounds?
9. Any cultural appropriation watchouts?
10. Do these stories communicate a variety of cultural experiences relevant to the audience?

These gut questions—and, indeed, the entire Playbook—can be used widely beyond the entertainment industry, in marketing, advertising, PR, communications, and even product development.