Is LinkedIn’s Algorithm Amplifying Gender Bias? A Critical Look at Visibility Claims

LinkedIn’s algorithm is a black box. We’ve all scratched our heads at its odd content choices. Inspirational quotes from unknown sectors? Daily routines of obscure CEOs? B2B sales insights derived from someone’s marital woes? It seems random, but maybe there is a pattern. Now, a more critical question surfaces: could LinkedIn’s algorithm favor men, boosting their posts while suppressing women’s?

For weeks, female LinkedIn users have been experimenting. Frustrated with low engagement, some switched their profiles to male, sometimes even altering names and using AI-generated photos. The reported result? A jump in visibility.
>

This trend arguably started with Cindy Gallop and Jane Evans. They asked two men, Matt Lawton and Stephen McGinnis, to post the same content simultaneously. Gallop and Evans had a larger audience, yet Lawton and McGinnis saw better engagement. Was the algorithm giving their male counterparts a leg up?

Since then, others have put the algorithm to the test. Megan Cornish, a mental health strategist, saw a 399% increase in post views after changing her gender on LinkedIn for a week. Here’s where things get interesting. Cornish didn’t just change her gender marker. She also used ChatGPT to rewrite her profile with “agentic” language – achievement-oriented buzzwords that you might find on a motivational speaker’s website.

She also rewrote underperforming posts in a similarly “bro-coded” style. In a twist, a LinkedIn post about the experiment, written in her own voice, didn’t perform as well as the agentic content. “I was confused, then it hit me: I used my own voice, not a fake male voice, for this post,” she wrote.
>

Felice Ayling, founder of Loud Women, heard similar stories about dwindling online visibility. When a fellow user mentioned changing her gender to improve results, Ayling followed suit. After forgetting she’d even done it, she checked her stats and discovered her impressions had more than doubled. She acknowledges the limitations, calling it “in no way scientific,” but believes a pattern exists that needs examination.

Nicole Ratcliffe, The Workplace Sleep Coach, also saw a drop in LinkedIn visibility. Posts with decent engagement reached only a fraction of her usual audience. Inspired by others, she spent 48 hours as “Nick,” complete with a male name and photo.

While her topics and language remained consistent, Ratcliffe observed that “Nick’s” reach expanded even with low engagement. A follower even reported only seeing her content after the gender switch. This raises serious concerns. Are women forced to mimic men and use “bro-coded” language to be seen? Do we have to trade authenticity for visibility?

Interestingly, race might influence these outcomes. Cass Cooper, a Black writer, experienced a decline in engagement after changing her gender. She suggested she was entering “a category that platforms and society have historically coded as less trustworthy,” highlighting the intersectional nature of bias.

LinkedIn addressed these claims in a blog post. Sakshi Jain stated that the algorithm doesn’t use demographic data like age, race, or gender to determine content visibility. The algorithm considers “hundreds of other signals,” including profile information, network connections, and activity. Jain added that anecdotal evidence of unequal reach “doesn’t automatically imply unfair treatment or bias.”

A LinkedIn spokesperson reiterated this, saying gender is not a ranking signal. They claim to regularly evaluate their systems for gender-related disparities. It’s worth acknowledging that the platform’s user base is predominantly male (57%). The algorithm could prioritize content aligned with male-typical behaviors, unintentionally marginalizing women’s voices.

Ayling suggests AI is learning and reinforcing real-world sexism. Algorithms learn from the tone, language, and content that gets rewarded, which could inadvertently favor masculine styles. She wants LinkedIn to investigate ways to give women a fairer shot at growing their businesses. Otherwise, she fears a potential exodus of female professionals, resulting in a less diverse and less valuable platform.

Ratcliffe plans to maintain her authentic voice on LinkedIn, despite this issue, because her profession calls for “honesty, emotional truth and lived experience.” Still, she believes the platform can’t be fully relied upon to circulate important content. More algorithm transparency is critical. As she puts it, “Visibility affects opportunity, and opportunity shapes the world our daughters grow up in.”

Read Original

Keywords: