Updated: Sep 20, 2021
I'm on Facebook, whining to my friend about a fight with my husband through their IM.
In my typical “melodramatic,” my husband claims, but I call “passionate” fashion, I'm furiously typing how much of a jerk he is, how often we fight, basically questioning if the sanctity of marriage is sustainable.
My friend is empathetic, and reminds me that I am always welcome to visit her in Britain, especially if I decide to leave my husband. She'll help me kick off my newly single life with a bang!
I have no intention of leaving my husband. I am sounding off, as women so often do to other women, about our marriages. And my friend knows this. She also knows her offer is giving me ground, a place to land. She is kind in the extreme, and I am honored to know her.
Unfortunately, Facebook is not so kind. They are collecting the IM exchange between me and my friend. Their data science team then runs Natural Language Processing (NLP) and Machine Learning algorithms on our text exchange. They categorize and classify our exchange to 'improve the engagement and response' to their advertiser's campaigns, aka “Personalized Targeting,” in marketing speak. They also sell our data to their advertisers, and you agree to let them do this when you download their app to your mobile, or sign up on your PC to engage with their platform.
In simple terms, personalized targeting means collecting as much data on an individual as possible, then correlating that data with people having the same or similar characteristics and behavior. With enough data, patterns of our behavior emerge. Facebook, Instagram, Twitter, and every other social network, and most every site we visit puts “cookies” on our devices, essentially tracking where we visit and what we do online. And they use that data against us, manipulating us to buy, try or subscribe to whatever they are selling through “Recommendation 'engines,'” aka algorithms.
RME, right? Another bullshit conspiracy theory.
What is bullshit is you not knowing any of this is happening, and even if it is, you believe it does not effect you.
In fact, “Recommendation” is selling you much more than products and services. Rec engines are telling you how to think. And you don't consciously know it. En mass, we are buying into too much of what is being sold to us online, aka “fake news.” Individually, digital marketing is now working at convincing me my life would be better if I divorced because 'the love of my life' is out there just waiting for me to swipe right.
So, after whining to my friend I'm off Facebook and on to my day. I social media market in the mornings for my books and entrepreneurial workshops before I start my day gig. I go to Instagram. My feed is filled with dating site ads, and many are, “Here is your second chance at love,” type campaigns targeting divorced women, or women considering divorce. Or women having a rough patch in their marriage. The ads keep flashing images of happy couples and my heart sinks a little. Instagram is owned by Facebook, in case you didn't already know.
For the next few weeks, many of the text ads in my feed on Gmail are for divorce lawyers in my area. Some claim cheap and easy 'arbitration' divorces. Sure, I've considered divorce. Any couple married for 25 yrs likely has. And with my husband and I still at odds, and feeling lonely, and craving intimacy, these ads present an option to staying married to a man that at the moment I'm not getting along with.
Over the course of the next month or more, usually way more, I get slammed with ads, essentially recommendations for dating sites and divorce lawyers wherever I go online.
The more I whine to my friends through IM, even email about my marriage, the longer I'll get slammed with dating and divorce site ads. The idea of dissolving my marriage, once inconceivable, which forced me to work it out with my husband, is now not only an option, but a promising one at that.
Facebook, Google, all of them justifying their use of Recommendation with marketing like, “more relevant and enjoyable ad 'experiences,' delivering content YOU want.”
Seriously? I don't want to divorce, so why is their Rec engine suggesting I should through the ad content they are continuously serving me?
Simple answer: so their advertisers—the dating sites and divorce lawyers—SELL more.
Conspiracy theory? Bullshit! The conspiracy is that you don't know, or even care this is happening to YOU too.
“Collaborative Filtering” is a form of Machine Learning that correlates your characteristics/behavior with people 'similar' to you. With enough correlations (data), computers can now see patterns in individual behavior. In other words, if I buy X, and your data correlates with mine, you will likely buy X too (see slide deck below). Advertisers use this data to slam you with ads their algorithms 'think' that you'll respond to.
Machine Learning and Natural Language algorithms are behind most everything we do online now. They are the gun on the table. But Recommendation engines are picking that gun up, targeting us, and pulling the trigger. Quite literally. Psychology 101: “Primers” (like the endless dating and easy divorce ads) and “Triggers” (like finding 'the love of my life' when I'm feeling lost and alone) are what motivate us to take any action.
The Recommendation algorithms behind Google, Facebook, Instagram, Twitter, YouTube...etc., are creating a FEEDBACK LOOP, showing you ONLY what you've expressed interest in, or someone like you has engaged with—IMed, emailed, searched, clicked, tried, bought or subscribed. You will rarely, if ever now, get a differing point of view, or opinion other than your own when you engage online. The Recommendation Effect is shifting politics—the global resurgence of right-wing nationalism influencing elections, and the predominant reason ex-president Trump was elected in 2016, and narrowly lost in 2020.
Conspiracy theory, my ass. The Machiavellian effects of Rec engines are real. And it is dangerous—socially irresponsible—to let social media platforms continue to collect our data and misuse it for profit at humanities expense.
Head's up to all you marketers out there: SEO based on relevance, link-backs, or even number of unique hits, is, well, gone for the most part. Returns on Google are now based on Recommendation—what Google's algorithms 'think' you want to see based on your past behavior and behavior of other's similar to yours. Google's “organic search” returns are also based on how much money they make from any given advertiser. The more they make from, say, Progressive Insurance Corp, the more likely Progressive will be close to #1 in their search returns. To get the broadest range of returns from any query now requires going to multiple search sites—DuckDuckGo, Bing, Yahoo and others.