Most Data Science Job Ads are Targeted to Young People - Why?
If you are on Facebook, you've probably seen their ads to attract candidates to data science positions. You've probably seen ads featuring many young women working at Facebook in data science roles, and while I think this is great, it is a misrepresentation of their workforce (70% are males, in engineering positions.)
What is surprising though, is that you would think these ads would be targeted to people who can relate to the picture featured in the ad. Yet, old white males see the same job ads with the same people being featured, than (say) young Indian females.
Here I discuss how Facebook's advertising for job ads (and it applies to many companies, not just Facebook, and also for ads about data science degrees, not just job ads) could be made more efficient, by creating fewer disappointed applicants, and not having Facebook go through tons of irrelevant resumes being submitted. We all think that theoretically, these resumes are automatically processed by artificial intelligence algorithms that are cheaper than human beings, but in reality, to this day, the opposite is true. It means that targeting the wrong people via these ads, or failure to attract the right people for the job in question, costs real money to Facebook -- and to the applicants wasting their time.
Facebook's business model is all about highly targeted advertising solutions for advertisers, and it has hundreds of data scientists to make sure that it is done right, especially when they promote its own campaigns on its own Facebook platform. In some ways, they look like the proverbial dentist who fix everyone's teeth, yet have the worst teeth.
Example of picture in a Facebook ad, to attract data scientist candidates
So, what kind or people (or other things) should Facebook use, in job ads pictures? Here is my take on this:
Rotate ads (Facebook does) showing a variety of people across multiple ads, changing daily (men, women, people of any age and race -- Facebook does not) to the same member, regardless of his/her demographics.
Show ads featuring African American people working for Facebook, to African American people (the same applies to women or old people.)
Show a big team featuring all types of people.
Do not show people -- instead show a Facebook cubicle, the Facebook restaurant, the nice weather in Palo Alto, or maybe even a smart dog with glasses seemingly solving a complex data science problem at Facebook.
Do not show job ads at all to people you are not interested in (in my case, all job ads that I see on my Facebook page are for junior positions, so this is a waste of advertising space.)
The issue might be with their legal team, forcing Facebook to show ads featuring minority people (except people 50+ years old), to comply with some government regulations, unfortunately resulting in advertising inefficiency. More on this in my upcoming article entitled The War Between Your Legal Team, Your HR team, and Your Data Science Team.
Tip to get a job interview at Facebook: Comment these Facebook posts featuring internal job ads, make sure you have a great Facebook and LinkedIn profiles, and that you never post stuff that is too controversial on Facebook. I was actually contacted by one of their recruiters, after adding a comment (carefully crafted) to one of these ads.
And it would have been for a great (senior) position that I would have enjoyed, and for which I was fully qualified. These ads typically receive 200 comments so you won't be the only one, but most of them (such as Why do I see this ad, I don't live in the Bay Area?) won't attract the attention of the recruiters. It is easy to outshine these commentators, and great comments are automatically featured at the top -- so recruiters are more likely to see them. And if you are a guy and see only women pictured in the job ad served to you (or the other way around), don't say anything about it -- it is a big taboo.
Some opinions expressed in this article may be those of a guest author and not necessarily Analytikus. Staff authors are listed https://www.datasciencecentral.com/profiles/blogs/m