Top

Would you like to:

TALK TO SALES? JOIN OUR TEAM?

Datawatch: from data science to dating science – how analytics can help you find love

In the latest in our series of blog exploring the most interesting analytics use cases, we examine the role data science plays in matters of the heart.

There’s a great story in an episode of NPR’s This American Life podcast titled ‘No Coincidence, No Story’. The tale focuses on a recently engaged couple, who while flicking through family photos discovered that the bride-to-be’s mother and groom’s late father were once engaged to be wed themselves, some 50 years ago on the other side of the world. The relationship ended long before the younger couple were born, and they may never had known had they not looked through that photo album. 

When it comes to matters of the heart, stories like this aren’t uncommon. You can find numerous examples of these seemingly otherworldly coincidences on the internet, which begs the question: what really brings two people together? Is it fate or some form of cosmic intervention, or something entirely more tangible and scientific? From die-hard romantics to grounded realists, you’ll find advocates in both camps, but what makes two people attracted to each other, and what makes that attraction lead to love, is one of the world’s enduring mysteries. 

Today, with the rise of online dating, data analytics is playing a huge role in trying to understand and simplify that process.

In the United States alone, there are 44.2 million online daters, and that number is on the rise. In 2015, dating app revenue in the US was $1.69 billion – by 2020 that figure had grown to $3.08 billion. And, according to a 2019 Stanford study, people are now more likely to find a relationship through online dating than any other method. 

With more and more people turning to apps and websites to find companionship, it’s down to the providers of these services to try and metricise the intangible and develop a science around attraction. Here’s how analytics is being used to do just that.

An inexact science

The problem with trying to use algorithms to find the perfect partner is that there are numerous factors that draw people to one another; a cocktail of physical attraction, common interests, similar moral frameworks and macro-economic factors. But what adds to that complexity is that every individual prioritises these things differently.

While some dating apps like Tinder simply match people with their desired gender and age groups within a certain distance, others are leaning more heavily on algorithms, machine learning and other techniques to try and solve this conundrum. 

Websites like Match.com and eHarmony use their own in-depth questionnaires to try and gain as clear a picture as possible of who their customers are and what it is they want. But the data work doesn’t stop there. As well as plugging this data into their algorithms, these websites analyse the behaviour of users on their sites, looking at the profiles they visit and how they match up with what they claim to want. 

This data is then further augmented by social media platforms, credit ratings, online shopping habits and a host of other factors to try and build a complete profile of compatibility. 

With this kind of data-driven matchmaking, there’s a very real opportunity to to delve into the essence of a person and create matches based on information people may not have thought to share themselves. By collating data from music and video streaming platforms, video game purchases, Kindle wishlists and other such sources, the chances of quickly finding people with common interest increases exponentially.

This is where data and analytics excels, in helping organisations to know their customers better than they know themselves. After all, when it comes to romance there is often a disparity in what people say they want and the people they ultimately fall for. With enough of the right information, dating sites are getting closer and closer to being able to uncover these known unknowns. 

The future of online dating

Online dating is a relatively new phenomenon, and as you might expect the algorithms and processes that support platforms and apps are improving day by day. 

For example, eHarmony’s matching system was originally built on a relational database management system, and its algorithm took more than two weeks to execute. Today, a more modern suite of data tools running on MongoDB can execute these algorithms in under 12 hours. 

This begs the question; if we’re still in the early stages of online matching, what will the future of this technology look like?

Two relatively new technologies that are already being implemented are deep learning and facial recognition. Dating site Badoo is using these tools to try and ascertain the physical characteristics people find attractive and locate similar users within their database based on that information.  

The UK dating app Loveflutter is another good example of emerging technology being put to the test. Using AI, the company matches people based on personality traits it identifies from their tweets – and is even working on coaching users through meeting online after analysing their chats.

As with any analytics use cases, the more available data there is, the more likely it is that these algorithms can match people that are genuinely compatible. And these algorithms could even be tailored depending on the traits different individuals prioritise in a prospective partner.  

It is of course too much to expect this technology to find the perfect match for everyone, but it can provide much better initial indications of whether or not you’re likely to be compatible with someone – something that years ago would have to be discovered over time. 

In essence, analytics can help shift the initial frogs-to-princes ratio, resulting in fewer bad dates and more good ones. The rest? Well, that might just come down to fate. 

Analytics at The Smart Cube

Here at The Smart Cube, we offer bespoke, end-to-end analytics capabilities, from data engineering through to reporting and visualisation, and advanced analytics. 

To read about some of the ways we’re helping our clients, or to learn how we can help you achieve your own business goals, visit here.