Summit Takeaways: What’s Diversity Got to Do With Analytics?

Lyndsay Wise
June 13, 2019

Words From the Wise

I was privileged to be invited, once again, to participate in our second annual Women in Tech panel discussion, moderated by Melissa Treier, vice president of Product Sales at Information Builders. Right off the bat, fellow panelist Karen Lopez (@datachick) drove home an “a-ha” moment that continued to be a theme throughout our discussion. She said that focusing on diversity alone "sounds like a quota-hitting exercise. What really matters is inclusion."

She illustrated a few of the practices that inadvertently exclude people, such as when companies are screening prospective developers and some recruiters look at resumes and decide whom they are willing to interview. They may be checking to see what contributions a candidate has made to the developer community, without taking into account their family responsibilities, volunteering activities, community outreach, and so on. The distribution of responsibilities in relationships generally accounts for some people having more free time than others, and more often than not, women are the caretakers and don’t always have the same amount of free time to blog, contribute to open-source development projects, or give back to their professional community. So they’re at a disadvantage on paper, before they even get in the door.

Information Builders' CEO Frank J. Vella kicked off the event to show his commitment to women in tech and commented on the challenge of inclusion. He discussed the “subconscious bias” held by many men in positions of power that also leads to a lack of diversity and inclusion, and proposed ways we can work as an organization to change the system.

So as we seek to leverage the promise of AI and broader data science within our organizations, we must ask ourselves a critical question: How do we support a culture of change?

Beware: There’s Bias in Your Data

Bias is everywhere, even in our artificial intelligence (AI) models. Dan Vesset, the group vice president of IDC's Analytics and Information Management, talked about Amazon’s AI hiring models, which showed their bias against hiring women in testing. So not only does the hiring practice require a lack of bias, but so does the model itself.

Girl, we’ve got our work cut out for us!

In many cases, organizations develop models believing AI technology will provide better insights, not realizing the potential flaws in augmented analytics due to the inherent biases in data generated by human beings. But it isn’t only about hiring practices and discrimination: the reality is that bias actually goes back to opportunities that we had or didn’t have access to, long before we first entered the workforce. In episodes 4 and 5 of his "Revisionist History" podcast series, Malcolm Gladwell discusses inequity in education, highlighting the fact that we do not have equal opportunity in today’s society, and that this lack of equal opportunity and societal bias begins way before we are considered for positions in the tech industry.

Time for Change

In my presentation “Aligning an Organization’s Data Strategy and Analytics Strategy,” with Information Builders' CMO Michael Corcoran, our discussion hit on the importance of change management. If you are trying to embed data analytics throughout the enterprise and need business and IT collaboration, but don’t have a seat at the C-suite, how do you change the culture and align stakeholders?

Change management is indeed required for a successful data management and analytics strategy, but the reality is that change is required on a broader scale. Think about it: if we address diversity and inclusion effectively, we will benefit from better analytics and insights. Being more inclusive helps to ensure that the AI models we build are less likely to be developed with the level of bias we have today. So instead of looking at women in tech, diversity, and inclusion as HR initiatives, we need to realize that the more diverse our mind share is, the more likely we are to become more competitive and take advantage of the augmented insights and broader data access that supports overall business success.

So as we seek to leverage the promise of AI and broader data science within our organizations, we must ask ourselves a critical question: How do we support a culture of change?