Does LinkedIn’s algorithm promote male profiles over female ones?
That’s apparently what several users have discovered by conducting their own impromptu experiments on the app, where women convert their profiles to male profile photos and names and then post the exact same content as female users, to test the results.
And some users have reportedly seen big differences, with up to 700% more impressions on the same posts shared as a male profile versus under a female name and identity.
Could that be true? Could there actually be an element in the LinkedIn algorithm, intended or not, that actively encourages posts from male profiles in the app?
Based on the number of messages under the #wearthepants hashtag there seems to be something going on in the app, so much so that LinkedIn now has it responded to the controversyand explained that the user’s gender is not an algorithmic factor.
As explained by LinkedIn Sakshi Jain:
“Our algorithm and AI systems do not use demographic information (such as age, race or gender) as a signal to determine the visibility of content, profile, or posts in the feed. Our product and engineering teams tested a number of these messages and comparisons, and while different messages received different levels of engagement, we found that their distribution was not affected by gender, pronouns, or other demographic information.”
So what’s going on then? Why do users get more reach when they post as men, instead of sharing the same or similar posts as women on the app?
Jain says there are many factors that play a role in reach, and it’s difficult to give a simple answer to why one post gets more impressions than another.
“A side-by-side snapshot of your own feed updates that are not perfectly representative or equally sized does not automatically imply unfair treatment or bias. Additionally, we see that the amount of content created daily on LinkedIn has grown rapidly over the past year, which means more competition for attention, but also more opportunities for creators and viewers alike.”
That’s a bit of a vague response, but essentially Jain is saying that many things, from the time you post, to the users who are active and see it, will dictate greater reach and impressions.
But it’s not gender, or any other demographic setting, that determines this. At least, not from LinkedIn’s perspective.
Another consideration could be the inherent bias of LinkedIn users, who are more likely to engage with a message from a man than a woman. These tests don’t take this possibility into account, but essentially it could be that LinkedIn users are more likely to respond to a man’s message when they see it in the feed.
I don’t know how you correct that, but it might be a consideration to take into account.
As for LinkedIn, Jain further notes that LinkedIn has internal testing to ensure that no one is “systematically ranked lower than anyone else,” to maximize opportunities, while also testing:
“…whether the feed quality for one target group is systematically worse than for the other, for example if women see more irrelevant feed items than men.”
Although the fact that LinkedIn is testing for this would suggest that it has settings related to male and female users, and that it is something that LinkedIn measures, at least to some extent.
That doesn’t mean that LinkedIn weighs messages from one group or another differently, but the fact that LinkedIn measures this experience also implies that LinkedIn could change its algorithm to influence the reach of messages from one group versus another, if it chooses to do so.
I don’t know, it seems like a strange point to make in this context, but essentially LinkedIn is saying that it has absolutely no weight in its system that gives female users less reach than men in the feed.
And of course that shouldn’t happen, even though LinkedIn has been working for years to maximize economic opportunities for all users in the app.
So I would expect LinkedIn to be more attuned to this, which goes back to bias testing.
It will be interesting to see if more users continue to raise these concerns, but LinkedIn says there is no gender bias within its systems.
#LinkedIn #denies #gender #bias #determining #reach #messages