Social Media and the Privacy Tug-O-War

By Lynae Cook, Social Media Strategist

In the late nineteenth century, America shuddered at the introduction of the postcard. The idea of practically inviting strangers to read your private mail was jarring. Just a few decades ago, we were horrified by the idea of tracking devices. Now we willingly carry at least one with daily, choosing the convenience over the endless Orwellian possibilities. The fears surrounding data privacy as it relates to artificial intelligence continue to dominate conversations surrounding marketing and advertising. We are presented the harsh reality, despite our fears, that only through sharing access to data can AI deliver the benefits that come from processing, analyzing, and anticipating the possibilities.

The present

AI has already integrated into our daily lives, delivering solutions for everything from innovations around health technology and energy solutions, to everyday interactions via digital and media. Is this reasonable? Probably. Did we give it permission? Yes. Did we know that? No. Do we expect it to continue to aid us, making our lives easier, while also resisting the potential negative outcomes? Absolutely.

How did we get here?

In a world where user agreements are pages long, and we (fairly enough) assume that if 2.19 billion other people have agreed to these user agreements, the rewards outweigh the risks. How could so many people willingly put themselves at risk to be tracked, hacked, or scammed? How could we have all the user agreement information at our fingertips and still be surprised when companies use the data we approved access to for their own bottom lines? Surely, if it’s publicly available, someone has read through and assessed it all, and enlightened media outlets, no? Let’s not be too certain, keeping in mind that the scandalous information about Enron was also publicly available prior to that scandal “breaking” (and was something that AI could have flagged if leveraged to assess information about the company).

While there are still a variety of ways the public needs education on data and AI, when it comes to social media, the questions are: What now? What next?

I’ll spare you the lengthy definition of AI terms, which, if you’re interested, you can find here. Instead, let’s focus on the matters at hand, with a few definitions inserted as needed.

Facebook is the perfect place to start, as it is the oldest form of social media we use, and the most widely adopted on a global scale. Facebook also owns Instagram and WhatsApp and has robust AI integrated into their platform, the most obvious existing as chatbots. Facebook’s image and video recognition capabilities are powered by machine learning, which is a branch of AI, and are often what companies are referring to when they discuss their AI capabilities.

What does this mean? Facebook now auto-classifies what’s happening in images and videos without human captions or tags. This is what Facebook uses to deliver the right content to uses across text, photos and videos, and influence how its ad product works, which as you can imagine, has significant implications for marketers trying to reach Facebook’s 2 billion users.

How do we succeed? It’s fair to say the news and entertainment industries often lead the way when it comes to innovation in social content — they are in the business of content, and platforms need quality content to drive value within their own communities. Due to this mutually symbiotic relationship, media companies form deep partnerships with the likes of Facebook, Twitter, etc. and work together for platform product launches to ensure there is quality content for users to explore when a new feature launches, setting the bar higher for everyone else. These deep partnerships give media companies insight in to what platforms prefer, meaning, what each platform’s AI will respond to best. Examples include the launch of Facebook Watch, Facebook Live, and Instagram Live.

A real-life example

Let’s look at the hit TV series none of us have heard of, SKAM. SKAM is a narrative built from the full breadth of Facebook content options: Facebook Watch, comments, texts, image uploads, and Facebook-owned Instagram content. SKAM is originally a Norwegian teen drama that was developed specifically because the network realized it was completely lacking viewership from the teenage demographic, a demographic also dwindling on Facebook.

The TV series launched with no advertising (so as to avoid the disaster of adults potentially recommending the show to their teenage children) and was launched specifically on social platforms. The show was such a hit that it recently came to the US as SKAM: Austin. This phenomenon may feel familiar to anyone who remembers the rise of The Real World and Jackass as teen favorites with a DIY aesthetic that truly stumped adults at the time. But in a world of endless options of interactions, imagine the teenage appeal of SKAM actors replying to your comments, following you, or producers closely monitoring conversations and adjusting plot-lines on the fly in response to user activity.

The major takeaway from SKAM is as follows: understanding human behavior as it pertains to your target audience is KEY to playing nicely with the AI that powers the social platforms we leverage for our clients. Why? It is in the platforms’ best interest to keep users there as long, engaged, and happy, as possible. Therefore, if they reward good behavior from brands, it is a better user experience for everyone.

In detail, what SKAM has done is notable with feeding algorithms:

• Audience insights — With months of research, the team behind SKAM identified exactly what would appeal to their audience and continues to learn about their audience and adjust accordingly.
• Content strategy — The producers of SKAM created a familiar art form (teen drama) through the lens of cross-channel content strategy. Owned GIF-uploads, Facebook Watch, Groups, Pages, Stories, comments, and messaging, and Instagram profiles, IGTV, Stories, comments, and messaging are all being leveraged in seemingly “real-time” to drive plotlines and the fanbase forward.
• Nimble creative development — Creative is developed and shared quickly and strategically across channels, content strategy has continued to adjust per market as platforms and audiences have shifted.
• Teams are integrated — Producers, directors, writers work in-tandem with social media strategists from Facebook to properly disburse content on the appropriate channels, engage with fans, and curate content appropriately for the Gen Z demographic.
• Content that adds value to the platform — This content is benefitting Facebook, so Facebook has dedicated representatives from Facebook Watch working with the SKAM team to get the best results disseminate content that resonates with the teens of today. The results is a variety of seemingly real-time content that includes video showing how teens are communicating using Facebook products.

This example is just one of many ways that success on AI-driven platforms is determined not only by the quality of content shared on them, but also their business priorities and a brand’s ability to add value for the audience it wants to reach.

Are there many things that have gone and will go wrong? Yes. Do I worry about how AI will impact visibility in to global problems as it silences content that it identifies “less than” preferred for platforms? Absolutely. In the meantime, between the hours of 9 and 5, we can work to better understand the people we’re talking to, how they behave where we’re talking to them, and share content that helps them find the right solutions… i.e. our clients.

Discussing the ways AI can go wrong isn’t fear-mongering, it’s responsibility in assessing risks. In the same way you’d want a plane manufacturer to assess all risks prior to taking off, we should do the same with the systems we use. I know AI can be scary, but ultimately, artificial intelligence and data will have the same normalized effect of postcards and “tracking devices” whether we embrace it today or not… because it already has.

Watch SKAM here, read more in this article in the New Yorker.

Learn how your #dog photos are training Instagram AI here.

Wildly valid concerns as stated in this opinion piece of Facebook’s algorithm.

Rational Ideas is the thought leadership hub for Rational, a digital agency and consultancy based in Seattle, WA.