Big data and the business of love

The use of big data by matchmaking service eHarmony is a thing to behold and highlights the potential power and reach of this massive technological change.

Relationships, love and romance are as old as time itself, but even they are not immune from technological change.

The rise of the internet has modified many of our behaviours. In terms of courtship, it’s triggered a boom in both online dating and matchmaking sites. 

While the ideas behind these websites – to match you with a better half - is rather basic, they are actually the product of a culmination of the latest business tech trends. 

For instance, have you ever considered how online dating giant eHarmony actually matches people?

While its marketing may point to its research and “compatibility formula”, the actual grunt work of this process is undertaken through the use of cloud computing and data analytics.

eHarmony has quietly been one of the big data trend’s most prominent advocates with the company using some form of analytics in its system since its founding in 2000.

So how does it all work?

Drop down boxes and questionares

Perhaps the best way to look at how eHarmony actually works, is to work through one couple’s experience with the system. 

About two years ago, a combination of “boredom” and a free trial offer led both Dean Rogut and his prospective partner Jodie into signing up to eHarmony.

They both recall being hit with a flurry of questions and “drop-down boxes” upon signing up for the site. 

“Religion, smoking [status], drinking [status], if you want any kids,” Jodie says listing off the top of her head the questions the system sought answers to.

Dean chimes in: “all forms of basic data.”

After entering in all the information, eHarmony’s system sets about the task of processing the data and matching with the site’s users. 

It’s a mammoth operation when you consider that each user has around 250 preferences attached to their account and that there are currently 1.5 million eHarmony users in Australia. 

Sparking a match

The system makes the easiest matches first, explains eHarmony’s international senior director of research, Dr Gian Gonzaga. 

It looks at what people do and don’t want in a prospective partner and rules them in or out of a shortlist. In Jodie’s case, a smoker would have been a deal breaker. 

From there, the system runs through all of the matches and begins ruling out people that don’t share commonalities that eHarmony has found to be inherent in good relationships. 

“For example, we found that shared spiritual values, shared passions, shared openness to experience are more likely to go with happy couples than with unhappy couples,” Gonzaga says. 

He adds that the system also takes factors that could be considered “conversation starters”, like similar interests or whether a match attended the same school or university.

After applying these filters eHarmony coughs up its first round of matches. It usually happens overnight. 

Jodie vividly remembers receiving only three initial matches. But why only three? After all of that data crunching there must have been more than just three people that were compatible with Jodie?

There was. But here lies another trick within eHarmony’s system: it limits your choices. 

“There was a study done a couple of years ago on speed dating, it showed that when people are given [the option of] 20 choices instead of 10 choices, people will want 20 choices, they think it’s better,” Gonzaga says. 

“But when they are faced with more choice, they actually do a worse job of figuring out who is the most attractive of that group and they are less likely to take action.” 

Gonzaga says that by only providing a few matches at a time, eHarmony manipulates the data to provide a better outcome for its users.

Machine learning 

eHarmony is also built upon a system of “machine learning” says Gonzaga. It watches its user’s every move and then adjusts its offering accordingly to better ensure success. 

For instance, if you’re only engaging with 30-year-old brunettes on the service, eHarmony’s system will match you with more 30-year-old brunettes that were previously determined to be compatible to your profile.

The deeper you dig into eHarmony’s system the more exponentially complex its data management task becomes.

Gonzaga says that while simply matching people based on strata is computationally easy, continually optimising those matches for every user requires a lot of processing power. 

As a result eHarmony has been quick to take up any innovations in both cloud computing and parallel processing, as they have become essential to it sustaining its service. 

Double dipping with the data

Big data has also played a role in optimising eHarmony’s business plan. 

eHarmony Australia’s managing director Jason Chuck says that the company employs a number of data analysts to churn through its users’ data and find ways to enhance its business offering. 

The company’s marketing, business decision and price points - as it is a subscription service - are all determined by the insights given from data emerging out of all of these system interactions. 

Average users, like Dean and Jodie, often operate on eHarmony without even considering the technology that underpins the site. 

“If I think about it, I guess there’s a lot of technology there, but I never thought about it... it just seemed to work,”  Dean says. 

eHarmony perhaps worked a little too well for Dean and Jodie. They were in each other’s first round of matches and never really got to see the full power of the company’s matchmaking system.

Since being matched in May 2008, the couple have been married for two years and now have nine-month-old son called Max.

They haven’t thought twice about the fact that their relationship was made possible by a machine. 

Jodie just sees eHarmony as a modern day take on the olden day “desperate and dateless” matchmaking balls. 

They both agree that the meeting people through sites like eHarmony is quickly becoming a norm.

Meaning that matchmaking through dig data is also becoming a norm. 

It’s an interesting point, considering that big data is one of the hottest trends in business at the moment. Using sophisticated data analytics is being touted as the most effective way to uncover consumer trends and is seen as a key pillar in the future of business decision-making. 

eHarmony may just be the first examples where we are seeing this trend come out of the realm of business and actually have a tangible, meaningful impact on people’s lives.