Stephanie Lampkin, Founder & CEO of Blendoor — All Raise’s Enterprise SaaS #WomanCrushWednesday
In May, we will focus on Enterprise SaaS and the brilliant investors and operators who bring their expertise to the industry. Our fourth feature this month is Stephanie Lampkin, Founder and CEO of Blendoor. All Raise’s ‘Woman Crush Wednesdays’ (#WCW) is a series where we highlight genius women who are funding or founding tech companies. Please come back to the All Raise Medium blog every Wednesday to find a new profile of an awe-inspiring female VC or founder.
Stephanie Lampkin, CEO of Blendoor, is on a mission to change the way companies think about people operations. Blendoor is taking a classic startup formula, leveraging technology to solve a big problem, to disrupt startups themselves and change how startups and enterprises hire. Studies have shown that large enterprises are biased in their hiring and haven’t evolved in decades. Researchers at Northwestern University, Harvard, and the Institute for Social Research in Norway did a meta-study of every available field experiment on hiring discrimination from 1989 through 2015 and found that at every step of the way, employers were more likely to proceed with white candidates. For example, white candidates receive on average 36% more callbacks than black candidates, and 24% more callbacks than Latinx candidates. Companies are biased to hire people that resemble the people that already work there, often to the detriment of company operations and performance. However, Stephanie Lampkin and her company Blendoor are looking to take the explicit and implicit bias out of hiring.
Blendoor uses machine learning and data analytics to do this. While other AI hiring models use data sets that result in a severely skewed candidate pipeline, Blendoor uses data sets that integrate performance metrics including manager reviews, quota achievements and retention data to better find a capable candidate. With Blendoor’s help, enterprises can find qualified candidates without losing them to unconscious bias.
Blendoor also releases a yearly “Blendscore” report that measures corporate diversity and inclusion based on demographics of leadership, retention, recruiting practices, bias and social impact initiatives. The 2017 scores can be found on their website and the 2018 score will be released at the end of May and include statistics on maternity leave and other benefits for mothers. Moving forward it will be a dynamic score, like a stock price, that will change as companies make changes.
Stephanie and I talked about how Blendoor helps enterprises find the best candidates and how she is uniquely qualified to tackle this challenge.
Q: What motivated you to start Blendoor?
A statement I sometimes give in my presentation is that no one is more incentivized to solve this problem than me- a 4’11” gay black woman. Like, fundamentally, the body that I was born into has strategically aligned me to solve this problem. But I’m a technologist. I see a gap and see a problem much the same way a lot of major inventors and innovators have and I am leveraging the resources and knowledge I have to start it.
Q: Why is it so important to leverage technology in our hiring?
A: We need better tools and resources. Unconscious bias is a challenge but you can’t manage what you don’t measure. A lot of HR organizations tend not to have very analytical people in those roles, despite people operations being pretty core to the business objectives. It is important to measure outcomes in order to optimize hiring and retention practices and that promotions and hiring are equitable across the board. The value that we bring is enabling them to have a dashboard that has data and analytics that show everything from top of funnel recruiting, hiring, promoting and compensating is equitable across the board and ensuring that outcomes are diverse and that your teams are reflective of the customers that you’re serving.
Q: Why is it important to take bias out of hiring?
Fundamentally, we suck at judging people. Enterprises tend to hire people who look like people they have hired before and they use procedures that haven’t changed. That’s why Amazon’s AI experiment didn’t work. They were using old data sets that had information from people who had already been hired. You’re baking in bias in the algorithm. Plus, we have seen the old way of recruiting doesn’t necessarily lead to better performance. For example, we see a strong bias towards candidates who have gone to top tier schools even though there are countless studies that show school does not have an effect on ability to execute in the job. 15% of Computer Science degrees were awarded to Black and Latino students but only 2% of the workforce at Google is Black and 3% is Latino. Their staff does not reflect the pipeline of talent. The further you are from a cisgender, white male in their 30s the harder it is to escape bias.
Q: How Blendoor go beyond the old data sets of companies?
We enable companies to hire based on merit and not molds. We are using performance data to train our algorithms rather than historical data sets. We can measure performance by just seeing how far along different types of candidates make it through every stage of the recruiting funnel. Once we get further integrated into an enterprise, we get access to their HR data. So we can actually see actual reviews and any sort of performance metrics, like sales, quotas, and retention data. We can use the data that companies are using for promotions and performance reviews. We can combine that with the signals from candidate resumes and LinkedIn to draw correlations between a candidate and a job.
Q: What are some areas where you see bias occur in hiring?
A: There are all of the obvious things you can think of like race, gender, University, age, etc. Unconscious bias is more likely to occur in any situation where people are resource constrained or time constrained. Some competitors sell based on the fact they can make hiring faster but the reality is that in order to combat bias, you do have to slow down your processes. You do have to be a bit more thoughtful and structured in the way that you do things. But sometimes that sentiment isn’t often aligned with the priorities of HR professionals.
Q: What are some best practices you’ve seen for enterprises wanting to improve their unconscious bias in hiring?
A: Blind resume reviews, blind coding exams, structured interviews (ensuring that all candidates are asked the same questions), diverse interviewers, demographic tracking (enabling the ability to identify where certain demographics of qualified applicants fall off in the recruiting funnel)
Q: How has Blendoor helped its customers? Do you have a specific case study?
A: We’ve enabled some customers to increase the yield of women hires by 2x and underrepresented minorities by 6x. But I’m unable to share specific customer details publicly at the moment.
Q: How does Blendoor help with candidate retention?
A: Our BlendScore review and analysis give us a bird’s-eye view into which and how companies are retaining diverse talent. Equipped with these insights and benchmarks, we are able to provide targeted recommendations for companies challenged with retaining certain demographics of high performing employees.
Q: Why should larger companies care about having a diverse workforce?
A: The next generation of the workforce take diversity, equity and inclusion very seriously. Top companies will lose their positioning as ideal places to work unless they effectively demonstrate that they care and are taking sufficient measures to maintain equity, diversity and inclusion. In addition to employer brand, there is sufficient evidence that diverse teams and companies with diverse leadership perform better. Thus, DE&I is both corporate social responsibility and a strategic business initiative.
Q: How does Blendscore work? How do you put it together? What are some of the interesting takeaways you’ve had from making it?
A: We intentionally make the methodology public; transparency is key. We analyze: Board of Directors and Executive Team diversity, US Workforce Diversity, benefits, development programs, recruiting partners, supplier diversity, social impact, and even other scores like Human Rights Campaign and Disability Equality Index. Unlike similar scores we want BlendScore to be dynamic, comprehensive and based on objective, quantifiable data. There are a ton of interesting takeaways, but most notably is the older/more mature the company is, the more likely it is to have women and underrepresented minorities in leadership. Also, in the three years we’ve been tracking the data, gender representation in leadership has grown substantially but underrepresented minority representation has remained stagnant.
