Algorithms have operated behind the scenes of our best-loved social platforms since their earliest days, but their side effects have only recently been scrutinized in light of Donald Trump’s rapid ascent to president-elect, and the UK’s surprising Brexit vote last summer.
These cultural moments taking place on either side of the Atlantic gained considerable momentum on social media, and in their wake there has been growing recognition of the role of Facebook et al. in bringing about their outcomes.
German Chancellor Angela Merkel recently weighed in on this debate , suggesting that Facebook and Google be made to disclose their secret algorithms, saying that users are entitled to know how the information they see is selected for them, and that it’s very likely this issue will come under the regulatory microscope.
Merkel and other critics of the algorithm trend are focusing on what in the tech world is called "filter bubbles" — personal online networks of content which continuously reinforce our individual bias and worldview.
"Filter bubble" describes the byproduct of social algorithms: These algorithms attempt to align with our own natural tendencies by selecting online content that is personally relevant to us, based on our previous clicks/shares/views. And these algorithms have become increasingly effective at that task.
The result is that our individual online experiences, from Google searches to Facebook feeds, are more filtered than ever. And content deemed too different to our audience profile is excluded — creating the "bubble effect" which prevents new points of view from challenging us. My view? Marketers should take a pin to that bubble.
Let me explain. This particular element of our online lives has fascinated us a while, so we delved into the issue here at our U.K.-based Unfolded Talks session, drawing on evolutionary science, psychology and various perspectives about well-being to understand the consequences of the filter bubble.
The result was that we agreed that social algorithms have not only magnified and automated our natural personal bias, but have made these bubbles completely invisible to us. That’s a process that should be a real concern for businesses — particularly marketers.
Freshly filtered news
What, where and how we consume content has never been more important, which is why online platforms have led the way in tailoring that information to users’ needs, based on their past clicks and views.
For social networks, this means suggesting news and connections that align with each user’s typical online activity. For search engines, it means that each user will find different search results than will the next guy for, say, "Tunisia," "climate change" or "best sports car."
In the short term, this scenario can be incredibly useful in an information-rich age. We are bombarded with so much information that surely it makes sense to filter out what isn’t relevant, right?
Maybe not, because when we take a long-term perspective, the problems become apparent. The events of this year are a sobering example. They show us that a long-term side effect of these bubbles has been the creation of extreme "us-versus-them" world views — demonstrated during the recent campaigns leading up to the U.S. election and Britain’s EU referendum.
Parroting isn’t personal, so challenge your audience.
Personalization is the buzzword behind these algorithms, as those in the marketing industry well know. Most ad-funded networks, from news outlets to social media, trade on their volume of audience insight, and it has become accepted conventional wisdom that "relevance" is "personal," and that "personal" is the goal advertisers should aim for.
The problem is, being personal is often interpreted as parroting overly similar content, products and services back to audiences that show related interests.
Instead, I’d like to suggest the value of a challenged audience. This is an audience which recognizes you amid the wave of look-alike content, understands that you stand for something different than do your competitors and is an audience willing to engage with your message, rather than to just "like" it.
That’s what I see as effective marketing, because it’s not reliant on simple relevance. Here’s how you can help remove your own audience members from their respective online ruts:
1. Understand the ‘present self’ versus the ‘ideal self.’
Netflix spotted a strange effect among its users switching from DVD rentals to streaming service: Those customers amassed a long queue of items that never got watched. The reason: They were lining up items they thought they’d like to watch in future, but never found themselves interested in watching them "right now."
This tendency is universal. We overestimate our future selves and, when it comes to Netflix, at least, we may queue up high art films, documentaries or classic movies that we know we "should" watch. Yet, in the moment, it’s easier to watch Hot Tub Time Machine instead of investing ourselves in something like Schindler’s List.
Considering this tendency, can advertisers encourage audiences to step back from easy gratification and embrace a world of alternatives?
Eli Pariser — CEO of Upworthy — described this phenomenon as balancing your "information vegetables" with your "information dessert." As marketers increasingly gain intimate knowledge of our online diets, encouraging us to vary them is simply a question of creativity.
2. Become a provocateur.
Brands can be hesitant to go against the flow, but look at what The Economist did: It recently reclaimed a huge swathe of readers by coming right out and directly challenging its audiences. Its campaign was made up of 60 pieces of creative content — each one asking its respective audience a tough question and questioning that audience’s assumptions about topics ranging from American police brutality to the European migrant crisis.
These provocations were placed directly into rival news sites of all kinds to disrupt audiences’ traditional browsing patterns and show them another point of view.
It was a simple but effective campaign which actually over-achieved on its targets: It delivered five times as many prospective subscribers as had been briefed (3.5 million people versus 650,000), with a return on campaign investment of 10:1.
Crucially, the mechanics of the campaign were fairly typical, and therefore easy to reproduce. Its provocative messaging was all that set it apart from a host of other digital campaigns.
3. Be creative with your platforms (or create a new one).
Spotify was in an uncomfortable position not so long ago. With Google Play, Pandora and more recently Apple Music to compete with, Spotify found itself in dire need of differentiation.
Here’s how it did that: Today, Spotify’s weekly Discover Playlists make millions of its users into fans every time it offers them a song they never would have found otherwise, thus throwing those listeners a musical curveball every so often.
The principle works in the bigger-picture sense, too. Toronto-based Million Short is an alternative search engine: It filters out up to a million of the first results that you’d find on the likes of Google and Bing, exposing its users to sites they might never have found before, and stripping away the obvious.
So, there it is: Spotify discovered the magic of serendipity. And Million Short proved (still does) how simple it can be to take users behind a curtain of standard content.
Follow their example! Take inspiration from their experience and give the people you’re targeting what they want — even if they don’t know it yet.
Yelena Gaufman heads up the strategic department at Fold7, an independent advertising agency in London. Originally from New York City, Gaufman joined Havas Worldwide London as its global planning director, before joining…