What referrals tell us about how networks work for people of color — Black, Latinx and Asian candidates
We are proud to share important and impactful findings from Atipica.
We analyzed referral data from the last year across all of our clients and we’ve found interesting findings that give us hope for thoughtful Diversity, Equity and Inclusion. Important to note — these findings reflect companies that are already actively invested in creating a more diverse workforce.
This is the first of a three part blog series.We will be releasing our findings weekly, centered on how referral programs affect different demographics. Be sure to check next week to read our next post on the experience of Asian candidates in the referral system.
Key finding: On average, referrals only make up 3% of all candidates, yet they account for 21% of all hires
Our concern: How can referrals be effective in empowering candidates of color, while not relying on current employee’s direct networks?
What we want to address: Let’s use data purposeful, inclusive AI and Atipica AI tools to include everyone.
Despite only making up 3% of the candidate pool, referrals make up a staggering 21% of all hires. Our data supports the evidence that nepotism and in network privilege still account for almost a quarter of all hires. Additionally, the referral pool is not evenly distributed by demographic.
Here’s what an average demographic breakdown of referrals looks like.
Nearly 50% of all referrals are white; 60% of all of those are men, meaning nearly 30% of all referrals are white men.
When a third of all referrals come from the same demographic background, the system is actively working against diversity and representation goals.
If you can’t get referred, your chances of being hired drop significantly, and unfortunately, the percentage of referrals who are Black or Latinx is significantly lower. This aligns with overall trends we see in low representation of Black and Latinx candidates across industries.
Key Finding: Black and Latinx candidates who are referred experience larger gains in hire rate than other groups.
Our concern: Black and Latinx job seekers who don’t have connections in the inside are less likely to ever be hired.
What we want to address: How can we elevate communities without access to network opportunities?
Everyone’s either been told or even said themselves that it’s “impossible to get a job without knowing someone”. While candidates are sometimes hired without having a connection, we found compelling evidence that for all candidates, their chance of being hired is dramatically higher if referred.
Our data supports the findings in this SHRM profiled study on employee referrals: a candidate who is referred has a drastically greater chance of getting hired than candidates coming in through other sources. It may even be the best way to get hired.
We found that any given candidate can expect to increase their chances of getting hired by 12x if referred.
This increase is not uniform across demographics. For example, the largest increase in hire rate is for Black candidates: when referred, the chances of being hired increased 16-fold (0.7% for regular candidates vs 11% for referrals).
We’ve known, from anecdotal evidence, that referrals are an excellent source of talent, but now we have the data to show that referrals are especially effective for Black and Latinx candidates, increasing their chance of getting hired by 16x and 14x, respectively.
At this moment, we want to acknowledge how problematic these findings are for Black and Latinx job seekers who want to enter the tech workforce without direct connections to those inside of it.
We know companies set goals on how diverse their employees are, not the diversity of their referrals. However, since 10% of Black and Latinx referrals get hired, we recommend increasing the number of inbound referrals of Black and Latinx candidates to make the most direct and impactful strides towards meeting those hiring goals.
Key Finding: Underrepresented talent is out there
Underrepresented talent is out there. Although Black and Latinx people make up 30% of America’s population and 14% of the total hires in our data, they only account for 10% of all referrals.
Given that research shows networks are homogeneous, especially for white Americans, it’s clear that, hiring more diverse teams will allow companies to access their diverse referral networks.
To increase the number of compatible candidates entering the pipeline from underrepresented backgrounds, look to your employees’ networks — not only who they know, but also what they support — specific alumni networks targeted at people of color, professional organizations and social graphs that enhance access to all.
In 2015, Intel spearheaded a referral program that doubled its referral bonus if the referral hired is from an underrepresented background. Since then, Intel has achieved each of their D&I goals, and companies like Accenture, Microsoft, and Johnson & Johnson have integrated similar programs into their own companies, seeing their own success.
Business Intelligence for the Changing Workforce
If you do not have a strong referral program, implement one! When referrals are made up of a diverse group of candidates is one way to ensure that people of all backgrounds are given an equitable chance of moving through the funnel.
Data shows that homogeneous social and professional networks exist across industries. In order to implement equitable referral programs, leaders will need to set up systems to center diversity and inclusion.
We recommend considering offer of increased referral bonuses for people of underrepresented backgrounds. At Atipica, our goal is to ensure your DEI investments are maximized. We offer services to facilitate the tough and necessary conversations for this to happen. We move beyond general training, by using your unique data we can incorporate behavioral data interventions that are proven to foster inclusion at work.
We know data drives action. Our team works with clients to get the strategic guidance your team and leadership needs now.
We believe in empowering people to engage the potential of their community graphs, not just rely on their direct networks. Find out more at www.atipicainc.com.
Methodology
Data Source on Demographics
- We honor self-reported data when available, yet our approach is to fill data gaps by using patent-pending modeling and proxies to identify trends that otherwise go unnoticed.
Average Percentage Breakdown of Referrals by Demographic across our Companies
- We found the demographic breakdown of all candidates at each company whose source field in our customers’ Applicant Tracking System (ATS) separately, and then took the average percentage of each demographic.
Percentage of Hires of Black and Latinx Background across our Companies
- Similar to the methodology above, we found the percentage of hires that were either Black or Latinx at each company, and then took the average.
Hire Rates of Referrals vs Non-Referrals
- For each company separately, we looked at all candidates whose source field in their ATS was “Referral” or something similar, and found the percentage of those that were hired.The same was done for all candidates whose source field was not “Referral”.
Increase from Referrals compared to Non-Referrals
- Increase in Hire Rate= Hire Rate for Referrals / Hire Rate for Non-Referrals
