Is LinkedIn Biased Against Women? What the Data Shows and What Senior Leaders Should Do About It.
UPDATED MAY 2026 - First Published 28 November 2025 - GENDER BIAS
LinkedIn insists there is no gender bias in its algorithm. Women's data is telling a very different story. And whether or not those two things can be reconciled, the lived experience of senior women on the platform demands a serious strategic response.
Late in 2025, a viral experiment called #WearthePants swept through LinkedIn. Women changed their gender settings to male, some also changed their names, and watched their reach spike by 700%, 1,300%, even 1,600%. The experiment was imperfect. Most women changed multiple variables simultaneously, making it impossible to isolate gender as the cause. LinkedIn's Head of Responsible AI responded swiftly and clearly: the platform's systems do not use demographic information as a ranking signal.
And yet. The pattern persists. Women's visibility on LinkedIn has been declining. Values-led posts are underperforming. Women of colour are reporting the opposite effect from the gender switch experiments. And the emotional labour of simply deciding whether to show up on any given day remains invisible in every conversation about reach and engagement metrics.
These two things can coexist: LinkedIn can be factually correct that it does not code gender as a ranking signal, and women can still experience gendered outcomes. That is not a contradiction. It is how implicit bias works.
The moment that named the pattern
While many women were running gender experiments, I published a post about two words spoken to a female journalist on Air Force One: "Quiet, piggy." It reached over 35,000 impressions in two days. Not because it was a LinkedIn strategy post. Because it named something every woman in a professional environment immediately recognised.
Being shut down. Being made small. Doing that instant calculation: do I speak up? Will it make this worse? What will this cost me? Whether we are talking about a president on a plane or an algorithm in a feed, the pattern is the same. Women's voices are easier to shut down than raise up. And the women who responded to that post were not reacting to a news story. They were recognising their own experience.
What the data actually shows
The #WearthePants experiments drew on documented reporting and platform-wide analytics. Geri Stengel's analysis of gender-switch experiments showed visibility spikes of 400% to 818%. The Guardian's investigation found increases of 1,300% to 1,600%, while noting that some women of colour saw the opposite effect. Controlled tests by Cindy Gallop and Jane Evans showed identical content outperforming when posted by men despite smaller followings.
LinkedIn's official position, reiterated by Sakshi Jain as Head of Responsible AI, is that gender is not used as a ranking signal. What the platform does acknowledge is that its new LLM-driven feed distributes content to clusters of users who behave similarly, rather than simply to a creator's existing network. If those clusters were built on historically male-dominated data, the system will naturally elevate the communication styles that data reflects.
As data ethics consultant Brandeis Marshall has noted, platforms are an intricate symphony of algorithms pulling multiple levers simultaneously. Changing a name or gender marker is just one lever. The bias, where it exists, may not be explicit. But explicit bias and experienced bias are not the same thing. And the women experiencing it are not imagining it.
What the new feed is actually rewarding
The shift from LinkedIn's old feed to its new LLM-driven model is significant. The old feed showed your posts to your network. The new feed shows your posts to clusters of people who behave like your existing audience. That means your followers are no longer your guaranteed audience. Your reach can shift dramatically when LinkedIn retests your account against different clusters. And if the system infers who you are from your writing style, tone, and topics, a miscategorisation can collapse your visibility silently, often without you realising for months.
What I have been tracking across client accounts since late 2025 is a clear pattern. Communication styles aligned with traditionally masculine norms, direct over warm, assertive over collaborative, declarative over reflective, are being rewarded more consistently. That does not mean women should perform a version of authority that is not theirs. It means the system has a default, and that default was not built with women at the centre.
What this costs women who are already calculating risk
The messages I receive from senior women about LinkedIn have a consistent texture. My reach has dropped and I do not know why. Posts that always worked suddenly do not. I am getting engagement but from the wrong audience. My practical posts perform better than my values-led ones. I am exhausted trying to think about LinkedIn all the time.
Underneath all of that is something the metrics do not capture. The emotional labour of being visible. The safety calculations. The tone policing. The fear of being too much. Women are not inconsistent on LinkedIn. They are calculating risk every time they show up. A system that adds to that calculation, even inadvertently, is not a neutral tool.
The rented land problem
This is the conversation I want senior leaders to have. Even if we agreed tomorrow that LinkedIn's algorithm is imperfect and subtly biased, the more important question is why so many brilliant women have built their entire professional visibility on a platform they do not own.
LinkedIn is privately owned. It was not designed with women's safety at the centre. It is fully capable of changing its rules overnight without warning or recourse. If your reputation, your opportunities, your inbound enquiries, your speaking invitations, and your sense of being known all live on one platform you do not control, that is not power. That is fragility.
I love LinkedIn. It changed my professional life and the professional lives of every client I work with. But I am firmly anti-dependence. The visibility strategy worth building is one that does not collapse when the feed shifts.
What a visibility ecosystem actually looks like
LinkedIn is the discovery layer. It is where the right people find you, form an initial view of your authority, and decide whether to look further. But it should not be the whole architecture.
LinkedIn for discovery and initial credibility signalling.
Email for depth, trust, and a direct relationship with your audience that no algorithm can interrupt.
Your website for authority, permanence, and a professional home you own entirely.
Press, speaking, and awards for amplification and third-party validation that no platform can remove.
In-person networking and community for the relationships that sustain everything else.
LinkedIn is your launchpad. It should not be your entire platform.
What I will not tell you to do
I will not tell you that you are imagining it. I will not tell you to rename yourself Luke. I will not tell you to perform a version of authority that is not yours or to masculinise your communication style to placate an algorithm.
What I will say is this. Name the bias. Work strategically with the reality we have. Build power in places LinkedIn cannot switch off. And refuse to disappear.
A final thought
The silencing is not always loud. It is being interrupted. Being glossed over. Watching less qualified people get more airtime. Or quieter still, an algorithm deciding your content is less relevant before you have even finished writing it. You cannot fix all of that. But you can decide you are not going to disappear. Your voice, your visibility, and your professional reputation deserve strategic care. Not scraps of your time at the end of the day. Not a platform you have no control over as your only foundation. Something more considered, more owned, and more built to last.
Laura Taylor is an executive LinkedIn strategist and ghostwriter working with senior women and C-suite leaders on visibility, reputation and authority. Based in Liverpool, working globally. ©️
