Skip to main content

Machine Learning service to empower a dating websiteDating Website

  • Client:
    Dating website, USA
  • Request:
    To increase customer loyalty and effective matches by using Machine Learning algorithm.
  • Result:
    30 million personalized offers per day and $3 million in cost-savings generated from user retention and effective marketing campaigns.
  • Technologies:
    • Big data
    • SPSS solutions
    • Hadoop
    • Apache Hive
    • Recommendation systems

Challenge

Our customer was an online dating service company from the USA (company name is under NDA). The client chose Cogniteq as it had experienced a very high user churn. The problem was the long time user spent in attempt to find a good match. 

The company set the business objective of increasing customer loyalty and user retention. The client aspired to maximize the value users get from using their online dating website.

Cogniteq team was assigned to execute this project due to vast experience in building portals and strong technical skills in Machine Learning. In addition, we had relevant experience in the business domain.

Our team was responsible for designing an algorithm for personalized selection of dating profiles. We further increasing the effectiveness of the client's marketing campaigns by exploiting Machine Learning technologies.

Solution

The use of Machine Learning was a key factor that influenced portal success. Thanks to the effective algorithm the solution has been changed in the way it functioned.

Questioning.
Now the website contains a detailed relationship-oriented questionnaire. The matching algorithm seeks to match people based on numerous parameters. The system recommends a user that special someone.

Selection criteria for a match.
People with the same habits build stronger relationships. In addition image analysis of the face in user photos adjust the algorithm so that the user is presented with profiles they would likely find appealing. 

Hence, the system connects alike-minded people who have the same habits or share similar interests. Moreover, the system analyzes user photos and selects those that are similar in composition.

The Machine Learning system explores the user's behavior i.e. how long he/she/they study various profiles, what type of appearance the user is more interested in, who they chats with, etc. The system learns from this experience in time helps users find compatible partners.

Improving digital marketing effectiveness.
Our experts integrated analytics to measure the effectiveness of marketing campaigns. The introduction of Machine Learning made it possible to use personalized advertising to promote website services.

Result

  • image
    Matchmaking algorithm offers thousands of matches every day.
  • image
    The company saves $3 million annually by increasing user retention and cutting marketing costs.