/Facebook’s Analysis on Relationships

Facebook’s Analysis on Relationships

Facebook ran a series of blog entries about relationships for Valentine’s Day in 2014. Here they are: the latest and greatest data on love from people who love data.

Love and Religion
The Age of Love
Flings or Lifetimes?
Looking for Love
The Formation of Love
When Love Goes Awry

Behaviour at the inception of a relationship

Relationships start with a period of courtship: on Facebook, messages are exchanged, profiles are visited, posts are shared on each other’s timelines. The above graph shows the average number of timeline posts exchanged between two people who are about to become a couple. We studied the group of people who changed their status from “Single” to “In a relationship” and also stated an anniversary date as the start of their relationship. During the 100 days before the relationship starts, we observe a slow but steady increase in the number of timeline posts shared between the future couple. When the relationship starts (“day 0”), posts begin to decrease. We observe a peak of 1.67 posts per day 12 days before the relationship begins, and a lowest point of 1.53 posts per day 85 days into the relationship. Presumably, couples decide to spend more time together, courtship is off, and online interactions give way to more interactions in the physical world.

Change in emotion level at the inception of a relationship

We used statistical methods to automatically analyze a set of aggregated, anonymized timeline interactions. For each timeline interaction, we counted the proportion of words expressing positive emotions (like “love”, “nice”, “happy”, etc.) minus the proportion of words expressing negative ones (like “hate”, “hurt”, “bad”, etc.). The above graph shows the proportion of positive over negative feelings being expressed in timeline posts before and after the beginning of a relationship. We observe a general increase after the relationship’s “day 0”, with a dramatic increase in days 0 and 1!

Chances of having a breakup as time passes

What determines the life of a relationship? By far the most important aspect is the relationship’s current age. The more you and your significant other have stayed together, the less likely you two are to split up. It’s easy to see why this might be. As the days since the two of you “made it official” (on Facebook) fly by, the relationship gets more opportunities to dissolve. Life circumstances and personal incompatibilities may lead to the relationship ending. Alternatively, maybe your couple is one of the lucky ones and your relationship chugs along to happily-ever-after.

With every month that passes, more and more shaky relationships end, leaving an ever-larger proportion of solid relationships around. This is exactly what we see in our data. We measured relationship starts and ends using changes in relationship statuses from “Single” to “In a Relationship with” and vice-versa. The above plot shows the monthly probability of a romantic relationship ending on Facebook, as a function of the relationship’s current duration, for all US users who have started relationships between January 2008 and December 2011. We only look into relationships that lasted at least 3 months on Facebook, and only consider only relationships where both individuals were at least 23 when they began the relationship, and only relationships in which individuals did not list their first relationship status as “married” (to try and capture something closer to true relationship starts, rather than the recording of long-existing marriages).

Behaviour at the demise of a relationship

The above graph represents the value of that weighted interactions score over time, relative to the person’s behavior before the separation. We observed a steady regime around the baseline before the day the relationship status changes, above by a discontinuity on that day with a +225% increase of the average volume of interactions which then gradually stabilize over the course of a week to levels higher to those observed pre-breakup.

This points towards people receiving support from their friends in times where they need it, whether it comes in the form of private messages, timeline posts or comments.

The team finds seasonal variation in break time, with the summer months (perhaps because they have a lot of young members) being the most popular time to rip out someone’s heart. The most important finding of this graph though? There’s a “small dip” in breakups around February. So, at the very least, you’re beleaguered relationship will likely have a nice Valentine’s Day.


  • If you and the other person start posting too much on each other’s wall, maybe you have started liking each other
  • Relationship keeps you relatively happier
  • The longer the relationship duration, the lesser the chances of breakup

An AI evangelist and a multi-disciplinary engineer. Loves to read business and psychology during leisure time. Connect with him any time on LinkedIn for a quick chat on AI!