Applied Analytics: Creating Serendipity


Our surroundings and experiences influence the way we perceive the world: how we think and act, the choices we make, and our general sense of taste. This multitude of individual tendencies creates what sociologist Pierre Bourdieu calls “habitus,” a cumulative pattern of the everyday that unconsciously informs our judgment toward selections of likeness.

The personalization of the digital space lies parallel to our “habitus,” as algorithms filter and strive to serve us relevance. As Mark Zuckerberg stated, “A squirrel dying in front of your house might be more relevant for your interests right now than people dying in Africa.” The filtered content, much like the aforementioned disregarded options, remains invisible – a process that we cannot influence.

 

The act of filtering

Let’s take the case of Spotify, which introduced a new approach to music discovery by acquiring Echo Nest, a music intelligence software. Echo Nest combines multiple filters to analyze music on an audio level and cultural level (using rhythm, tempo, and timbre for the former; and by reviews, ratings, and tweets for the latter). The former deepens our music discovery by giving us more of the same, while the latter broadens our music discovery. But again, the addition is filtered and thus limited.

 

Introducing serendipity

(Left) Digital content is often filtered based onpast patterns and connections, making discovery difficult.

(Right) But we can cultivate a state of “controlled serendipity” by introducing filters to intentionally introduce diversity into our digital experience.

 

So how can we make our digital experiences more diverse? Dismissing filters per se is not the solution; without them, we would be overburdened by abundance and too much choice would paralyze us. (Barry Schwartz’s notion of the “paradox of choice” explains that the more choices we have, the fewer choices we make.) That’s where serendipity comes into play (whose definition according to the Oxford Dictionaries is “the occurrence and development of events by chance in a happy or beneficial way”). Serendipity is important for diversifying tastes and (commercially speaking) driving users.

 

The digitization of serendipity

Serendipity in its purest form is not reproducible in the digital space. As soon as people start to “choose” serendipity (i.e. shuffle music playlists), their experiences can no longer be seen, strictly speaking, as serendipitous. Nevertheless, we believe that the idea behind serendipity is worth pursuing. While we cannot reproduce serendipity in its purest state, we might cultivate a state of “controlled serendipity.” Controlled serendipity can be a tool for making the bubble of algorithmic filters, in which we are currently browsing, permeable.

The challenge now is to start reframing our digital experiences in order to create a state of controlled serendipity. And there are numerous ways of doing this. One solution is to filter only a certain percentage of our digital experience based on relevance (e.g., 80 percent), while the balance (e.g., 20 percent) remains unfiltered and diverse.

Another example is algorithmic. Instead of only filtering according to relevance, different criteria – such as whether something is “challenging,” “important,” or represents “other points of view” – could be applied. These criteria can vary according to platform types and their purposes (i.e.,  search engines, social networks, news platforms, or e-commerce websites).

Regardless of strategy, we consider controlled serendipity to be important in ensuring that users are exposed to as many distinct experiences as possible. And we believe it to be commercially relevant in an overcrowded marketplace, where companies are keen to deliver unique and memorable user experiences.

As Sir Tim Berners-Lee so eloquently put it, “We need diversity of thought in the world to face the new challenges.”

Download Full PDF
Related Keywords
  • eBusiness & Channel Strategy
  • Big Data
  • Digital Experience
  • Modern Service Delivery
  • Personalization
  • Automotive
  • Communications
  • Consumer Products
  • Financial Services
  • Hospitality
  • Other Industries
  • Retail