by Sneha Jha Uses Analytics to Figure Out What Women Want

Sep 10, 20147 mins
AnalyticsBig DataBusiness

Do women prefer smokers, non-smokers, or social smokers? uses big data analytics to unearth that and more. 

Gone are the days when all-knowing aunts were pressed into service to seal suitable alliances for eligible bachelors and spinsters. When it comes to matchmaking, online is the way to go. With hordes of people taking the plunge into Internet-arranged marriages, India’s e-matrimony business is poised to grow to Rs 1,500 crore by 2017, according to ASSOCHAM. owns a huge chunk of the cyber matchmaking pie market. With brands,,,, and new business models like (a virtual relationship manager-based service, for those who find hunting partners online daunting or don’t have the time), the Chennai-based Internet company is nearly twice the size of all of its Indian competition put together. 

“We cater to a large number of online matrimonial search subscribers searching for soul mates in the virtual space. We have about 8,000 members registering every day on our website, and that number’s been swelling with every passing day. We have about 1,500 marriages reported in a day,” says J.K. Iyer, chief strategy and analytics officer, It has in subscribers in nearly 5,000 Indian cities, and NRIs from almost 150 nations. The Brand Trust Report 2014 ranked it as India’s most trusted matrimony brand.

As great as having all those members was, there was a flip side: It made it very hard for to understand and engage effectively with smaller groups with specific needs.

The Hitch

The bedrock of successful cyber matchmaking is organized communication; communication between and its members, and between members themselves.

One of its communication goals is to motivate members to complete their profiles. Another is telling members how to improve responses to a prospect.

To do this, interacts with its subscribers through multiple channels. These include a 2,000-plus telesales team, a customer support team that handles 3 lakh calls a day, its feet on the street, via SMS, and e-mailers, and using its website, its mobile site and 180-plus brick and mortar offices. 

But as grew, communications started to take place in silos. It was a knotty affair, for example, to pull out the communication history of a single customer across e-mails, telesales, SMS, customer support, feet on street, and retail stores, among others.

“Our communications are multi-channel and multi-wave. Not all our members will respond at the same time or respond to one channel. So, communications need to be multi-stage. If we now add other communication agenda, the communication agenda becomes very complex,” he says.

The second big ask was analytics. “We wanted to know who responded to our communications and campaigns, and which member sent out alliance requests to other members. We also wanted to garner deeper insights from who married who. It is obvious that such intelligence will be extremely useful in delivering our brand’s core promise: ‘Find someone who cares for what you love’, and, of course, the larger and ultimate goal of happy marriages,” says Iyer.

Analytics could also help tweak the profile form asks its users to fill.

Filling the profile form is critical point in the marriage hunt process. Subscribers sift through a host of profiles before selecting a suitable candidate. That makes profile information crucial. The profile form is quite comprehensive, comprising several mandatory and non-mandatory fields.

“We keep a some fields non-mandatory at the time of registration since several questions require time to fill. The ‘about me’ field, for example. The key is to understand which of the mandatory and non-mandatory fields are critical for alliance generation,” says Iyer. figured it could use big data analytics to generate intelligence to advise customers what fields have strong correlation with alliance interest generation and how they should fill them. That’s invaluable information from a member’s point of view.

The variety and volume of data is flooded with is vast. But there was a need to link the oceans of data in telesales, customer support, Web/mobile data, retail, feet-on-street, e-mail/SMS for integrated campaign management and analytics.

Iyer decided to use big data and analytics to help the online match maker analyze its vast troves of data and create meaningful campaigns. He should know: For about 15 years, Iyer has been teaching marketing analytics to consulting firms across the globe. 

Perfect Match’s big data analytics journey commenced about 18 months ago. Iyer started by creating an internal vision documents around analytics and campaign management. He then undertook internal pilots.

In the first quarter of 2013, he conceptualized big data analytics project. His first job was to zero in on a solution that met three criteria. He wanted a solution that offered the best integrated marketing communication suite in terms of its ability to integrate multiple campaign sources such as Web, telesales, e-mail, SMS, etcetera, and was capable of handling diverse data sources, and handle scale.

The solution also needed to include a robust analytics that could handle very large and diverse data types. His final criteria was a great implementation partner. He cherrypicked IBM.

Work quickly got underway and in the first phase, implemented the analytics solutions and deployed in April of 2014. Campaign management tools are now being implemented in phases across all channels.

Right Click

Clearly, the benefits are worth all the effort Iyer and his team put in. The project, says Iyer, achieved what it set out to do: Give the company insights into subscriber preferences.

“In a male subscriber’s profile, we have a field that captures information about his smoking habits with three options: Non-smoker, social smoker, and frequent smoker. The big data analytics tool has helped us re-evaluate our understanding of how women reacted to this information,” he says.

What big data analytics unearthed was that women preferred non-smokers, smokers, and then social smokers, in that order. It was a counterintuitive revelation. Until then, executives at believed that women preferred non-smokers, social smokers, and then smokers.

Iyer shares another example. “We analyzed volumes of text data and figured that women are most interested when men write about their hobbies–and least when they write about their expectations. These is extremely crucial intelligence about what should be included in profile descriptions,” he says.

The organization’s national presence and the diversity of its customer’s socio-economic profile calls for segmentation. This can ensure that its sales goals are better met through targeted campaigns and offers.

“Campaigns are not just about communication–but also about determining, through analytics, what product would suit them and when. Analytics also helps determine channel propensity,” says Iyer.

The online match maker is now equipped with a much better understanding of its members. Now it’s bettered armed to use the right products, platforms, and campaigns to engage with them.

“We have reduced three fields from the profile information form. Our alliance interests have shown early rises by at least 5 percent. Similarly, there are signs that sales have grown at least by 8 percent in the last two weeks of modifications and changes. These results are from very recent improvements. Over time, we expect most key metrics to move up significantly. That also means smoother member-to-member interactions and very good traction in alliance interests –all in all more happy marriages,” says Iyer.

Big data analytics project revealed interesting twists in women’s choices.