I’ve had the opportunity to be a product leader in both New York City and San Francisco and have called the bay area home for the last 1.5 years. I feel lucky to work with some of the smartest technologists in the world here and am constantly challenged to up my game. However, one thing I am sorely missing is more diverse product teams within silicon valley. I’ve seen that diverse problem solving leads to product innovation. Part of me took for granted how much easier it was to create and hire diverse teams in New York City in part due to my involvement in the distinct NYC melting pot culture. It still wasn’t a solved problem as not all our teams were diverse and NYC has it’s own technology diversity problem. However, It’s well documented how bad the diversity problem is here in the bay area, particularly in regards to underrepresented ethnic minorities, ageism, and gender biases.
When I define “diverse thinking” this includes not only things you are born with such as gender, race, sexuality and age but also experiences that people acquire over time such as working in another industry or in another country. If we want to build products and services that the whole world wants to use, diversity in thinking is a strategic competitive advantage and product organizations that get this right will build the next generation of meaningful experiences.
An experience I had a few years ago in the Image/technology industry in NYC continues to resonate with me. Below is a summary of how a group of people came together bringing their authentic selves & backgrounds to the table to create a great customer experience. We were trying to solve a customer problem — we knew that users were comparing images to one another in a variety of ways before making a choice of what to purchase, and we had a hypothesis that if we helped them make these comparisons more efficiently we could increase usage.
- One of the team members, a man in his late 40s who came from the photo industry was able to bring in a physical lightbox with magnifying glass to examine the photos for us to learn about. He talked about the goal of the photo editor during this process — comparing color, focus and other photo criteria quality.
- A mid-twenties front-end engineer from Puerto Rico came up with one solution, a way to preview photos on the site at a much larger resolution, allowing customers to get an in-depth view prior to making a purchase with a single click.
- A mid 30s female Chinese machine learning engineer from our search team came up with an algorithm for visually relevant images that could be seen from a photo detail page.
- A black woman in her 30’s suggested that the best place for comparison was within the shopping cart after users have finished with their browsing experience and have moved into ‘compare and purchase mode’ based on user research sessions.
- A late 20’s MBA, formerly in finance, but moved to product management with no technical background built a forecast to show how even minuscule features in previous A/B tests that stood in between the image preview and the purchase suppressed conversion, therefore opting for a different, more incremental recommendations feature.
All these problem solving frameworks were imperative to creating an awesome experience for customers. The team in the example above wasn’t hand-picked, it came together organically — lucky for us! What’s important is that we as an organization had a robust list of innovative solutions from a variety of perspectives (most of which found their way to the production site). In this sense, diversity is just good business because different problem solving frameworks lead to better products.
The anecdote above tells a great story, but there’s data to support this hypothesis: Scott Page and Lu Hong conducted studies in 2004 where they compared teams of “high performers” vs. teams of randomly selected individuals and found that diversity trumps ability. They looked at a variety of criteria, including IQ, and created problem solving experiments where groups of high IQ individuals with similar backgrounds were compared with a group of mixed IQ scores with diverse backgrounds. The diverse group with mixed IQ scores statistically won most of the time.
One of the reasons why diverse groups outperform groups with similar backgrounds are the different frameworks people use to solve problems. I had the opportunity to experience this first hand when switching from the music industry to the image industry. In music, there are lots of interesting ways that musicians manipulate audio, via filters, sliders, compression, etc. However, when I changed my career to the image industry, many of these techniques were considered innovative when applied to image manipulation. These problem solving frameworks weren’t innovative to me, these techniques were simply applying previous experiences to a new market, but herein lies the benefit of diverse thinking.
One can make the counter-argument that task-based work is faster to complete by individuals with a high IQ than individuals with average IQs. Therefore teams of people with similar (not diverse) backgrounds would be more efficient. However, the types of problems solved in product development are rarely task based, unless you work in a place where your team is building the same widgets over and over again. Most product development teams, particularly those working on feature development, are solving problems for the first time and the type of problems they are solving are complex, ie, a number of different approaches are required to solve the user problem. As such, the types of problems in product development are the types that benefit greatly from diverse thinking.
One of the aspects of great, diverse teams that Scott Page’s study above assumes is that communication within the teams in the experiments are equal — ideas are generated, reviewed and executed equally amongst all the team members. Team communication in the real world varies widely, which leads to another concept imperative to successful product teams: Innovation requires not only diverse thinking but also the ability to share ideas openly with each other, devoid of attack. In Charles Duhigg’s book “Smarter, Faster, Better” he references Google’s study of high performing teams and one of the key elements to performance is psychological safety. This safety is created through empathy for your teammates, specifically, people from diverse backgrounds feeling unafraid to share personal information or bad ideas. The combination of diverse thinking and psychological safety give teams the best chance at innovation. Inversely, diverse thinking without empathy for each other’s perspectives can lead to unhappiness amongst team members where the various problem solving approaches of the team clash and become inefficient.
So why is this topic so important for leaders in technology to debate? Technology is outpacing our ability as humans to adapt to and humanize it, and a lack of diversity in product development is exacerbating this problem. Let’s take one of my favorite Silicon Valley topics, artificial Intelligence (AI). I’m sold on AI, its time has come — we have the computational power and more importantly, we have the rich datasets to leverage AI to create value for humans in a way that’s been over-promised and under-delivered for almost 30 years. I was listening to an inspirational A16Z podcast “When Humanity meets AI” when I heard Fei-Fei Li (Head of AI at Stanford) mention that today, Artificial intelligence as a field is sorely lacking in diversity and “humanistic thinking.” It reminds me of how the web looked back in the early 1990’s before any designers got involved in its creation. It was clearly headed the right direction, but the language, appearance and overall experience wasn’t translating to a large diverse customer base. AI is here — from a technical perspective, but it’s not here from a diverse humanistic perspective. Some will claim that AI just needs a visionary breakthrough from a genius product leader to go mainstream. However, I’d be willing to bet that if we injected diverse thinking from a variety of different races, genders, culture, experience and age into the product development process of AI, we’d see innovation in the space much faster.
Every leader has an obligation to proactively work to make change for underrepresented groups and experiences. I also think that this is just good business and will lead to more innovation: The next phase of technology innovation will be fueled by product teams who are capable of empathizing with and understanding diverse problem solving frameworks. These teams will realize that one of the best ways to experience diverse problem solving frameworks is to be around people who are — by way of race, gender, age, sexuality or work experience — looking at the world through a fundamentally different lens than they are. Choose your team members wisely in both the hiring process and when mixing teams up internally. Focus on traits such as perseverance and variety of experience of a candidate over the Ivy league school or brand of previous company. Here’s a great post with some pragmatic ways to introduce diverse thinking into your team or organization.
Thanks Celsea Jenkins, Lyndsey Williams, Lindsey Scott and Mark Sherrill for the edits/thoughts.