AI Product Management: Breaking Gender Barriers in a Technological Landscape

Author: Shreya Thakur

In recent years, the role of the product manager has undergone a multitude of changes. The genesis of the product manager role can be traced back to the early 20th century. In 1931, Neil H. McElroy, a junior executive at P&G, wrote a famous memo proposing the idea of a "Brand Man" who would be responsible for a brand's success—including advertising, promotions, tracking sales, and product changes. This was the kernel of the concept of product management, which ushered in the need for skilled personnel overseeing the entire lifecycle of a product.

Cut to the 1980s - as the software industry boomed, companies like Microsoft and Intuit began applying product management principles from consumer goods to software products. Early software product managers, often coming from engineering backgrounds, served as a bridge between development teams and business stakeholders. As product managers started shouldering more crucial responsibilities about product vision, strategy and execution, product management became an established discipline with a growing body of knowledge and communities.

A Nebulous Role with Prevalent Gatekeeping

The role continues to evolve in tandem with the tectonic shifts ushered by technological advancements, especially in the realm of AI. A strong technical background is often considered a prerequisite for AI product management roles. Many companies have historically favored candidates with computer science or engineering degrees, data science expertise, and experience with machine learning for AI Product Manager (PM) positions. This has created an additional barrier for women, who are already underrepresented in those fields.

In a field where women are already underrepresented, female AI-based product managers face additional challenges navigating multi-stakeholder environments dominated by men. This issue can be more pronounced in the case of newcomers as they may struggle to have their voices heard, their perspectives valued, and their authority respected in the absence of more women in leadership roles to advocate for and support them.?

To see an AI product through from conception to launch, adoption, and scale requires great patience and persistence. A PM's role is a spectrum of tasks and activities that require asking the right questions, communicating their ideas effectively, convincing stakeholders of optimal outcomes across various options available, and involving them in the process of co-creating the products being built.

The ‘Wicked’ Side of Things

While women in the AI product management space face challenges in both the private and public sectors, the nature of issues in each milieu can differ. AI relies heavily on large amounts of high-quality, relevant data to train models and generate insights. Collecting, cleaning, and managing this data is a critical part of the AI product development process.?

Furthermore, developing AI systems requires constant iteration and specialized skills such as expertise in data science, machine learning, and specialized tools/infrastructure. The models need continuous training, testing, and updating as new data becomes available. The development process is also experimental and comes with risks of AI models perpetuating biases.

All of these factors make stakeholder involvement uniquely complex for AI products, especially in the public sector. AI systems can impact a wide range of stakeholders beyond direct users, including people affected by AI decisions. Building products for the government and public sector often involves navigating a more complex web of stakeholders wherein decision-making is seldom black or white, stakeholder buy-in is critical, and problems run the risk of being categorized as wicked.?

A wicked problem is a complex, challenging issue that is difficult to clearly define and has no straightforward solution. Wicked problems are characterized by interconnected factors, competing stakeholder interests, and potential solutions that may lead to unintended consequences in other areas.?

Often the goals of various stakeholders might be too divergent to align, or too incompatible to conform to the idea of building a viable product. The principles of AI system design, data infrastructure requirements, and AI ethics and policies may seem too disjointed. Product management units that operate within the research, non-profit, or public sector apparatuses have to curate a program to budget for AI skills, time, data assets and externalities, and orchestrate its execution while ensuring multiple stakeholders are on the same page. An AI product manager thus navigates a multitude of small yet critical decisions and dialogues daily, akin to a film director who strives to create a visually stunning movie based on a script.?

Overcoming Barriers in Stakeholder Management

A major challenge for women AI product managers in this journey is the lack of representation in decision-making discussions and AI strategy. Convincing stakeholders on priorities and problem-solving methods is tough, especially within traditional frameworks. Introducing a new perspective on responsible AI development can be sticky when you are the sole person at the table with a standout perception or if others need help to think divergently to see your viewpoint. These bottlenecks can be overcome by maintaining a balance between assertiveness and an acknowledgment of the thought processes of the other stakeholders.?

People management and communication skills are crucial for understanding the values and beliefs of your stakeholders or the individuals you interact with when building AI solutions for the public sector. By understanding their perspectives, and potential concerns around AI risks, you can tailor your approach to persuade them more effectively. For instance, if a stakeholder is worried about the AI model’s accuracy and fairness, you should provide a concrete mitigation strategy that matches their mindset. It's essential to invest time in understanding your stakeholders and improving your communication skills. The more you focus on these areas, the better prepared you'll be to handle challenges successfully.

Systems thinking skills are also a worthy ally, as they allow you to explore unexpected perspectives or alternative narratives that the individual stakeholder might miss. You can introduce a systems approach to steer the stakeholder toward your desired viewpoint.

The other indispensable ingredient in this recipe is data which can act as an impetus in such situations. Data can show a stakeholder the scale of a problem and help assess whether the proposed AI solution will be effective. Some problems may appear critical because they are one-time anomalies or accidental events. With sufficient data on the frequency or prevalence of such events, the need to allocate resources for an AI-driven solution can be evaluated accordingly with the stakeholders.

Repairing the Broken Rung: Strategies for Progress

Organizations have an equally vital role to play as women AI product managers strive to ensure that the AI product management landscape has robust diversity.

Women in AI product management roles must put themselves in situations that facilitate learning. This includes enrolling in relevant AI/ML courses, attending industry conferences on AI product innovation, and being up-to-date with industry literature. Companies can supplement this approach by offering stretch assignments and rotational programs that expose women to teams working on different AI solutions and use cases.

Approaching AI/ML teams with a growth mindset to leverage their knowledge and capabilities is crucial for building credibility. Actively participating in discussions around pertinent topics such as data pipelines, model architectures, and MLOps is a must and helps gain exposure.

To drive lasting change, supporting women enrolled in university STEM programs is crucial. Companies can provide improved internship opportunities, offering targeted mentoring and coaching as women prepare to transition into the workforce, and actively recruiting women for leadership roles in tech projects.

Note: This blog is made possible by the support of the American People through the United States Agency for International Development (USAID.) The contents of this blog are the sole responsibility of Wadhwani AI and do not necessarily reflect the views of USAID or the United States Government.