Sign up for our newsletter and get the latest big data news and analysis.

Tinder and Interana Find a Match

BigData use caseTinder uses behavioral analytics solution, Interana, to understand and explore how its tens of millions of users interact with the app. Generating massive volumes of event data from over 1.8 billion “swipes” per day and 26 million matches per day, Tinder is one of the most used social applications across the world. Gaining deep insights from its event data with Interana, Tinder is providing the best social discovery application to its users.

Enabling people all over the world to connect and create meaningful relationships requires an innovative solution that has wide appeal and keeps users engaged. In 2012, Tinder launched its game-changing social discovery application with tremendous success. Quickly rising to the top with tens of millions of users, Tinder accumulated a surging amount of data. Through all this success, Tinder was on the hunt to upgrade to the best behavioral analytics solution that could handle their massive scale with built-in behavior-based tools, provide self-service analytics so all departments could benefit, and be cost-effective to implement and maintain. When Tinder “swiped right” on Interana, it was an immediate match. For Tinder, Interana delivers a behavioral analytics solution for event data that:

  • Scales into the trillions of records without sacrificing performance to deliver insights in seconds rather than in hours.
  • Provides out-of-the-box behavioral tools for key types of analysis like conversion, engagement, retention, and root-cause analysis. The Tinder team has the pulse of the app at their fingertips.
  • Democratizes data with an end-to-end self-service experience that puts data at the center of decisions. Tinder is a data-driven organization with several departments like marketing and engineering analyzing data in real-time.
  • Delivers a streamlined, full-stack solution that is quick and easy to implement, and maintain. By combining the database, analytics layer, and visual front-end, everything works together seamlessly and does not require an army of people to maintain or use the solution.

Tinder ImageWhat to Do When You’re Drowning in Data

In 2012, Tinder lit up the tech scene with its matchmaking application, quickly surging past the industry’s initial growth expectations. Tinder’s success is comparable to other applications like Snapchat and LinkedIn with regards to its active use and popularity. Today, Tinder operates in 196 countries, whose users generate an average of 1.8 billion swipes and create 26 million matches per day.

In Tinder’s surging data, there is a wealth of insights to help the company continue to improve the app to attract new users, deepen engagement, and retain users longer. Since Tinder’s successful launch and its expansion into new regions, it constantly looks to its data to see how people interact with its app. Things like:

  • When and why usage increases.
  • How many times a user swipes right or left, and why.
  • Discovering new social norms in different countries.

Very early, Tinder understood that they needed insights from their data to continue developing the right strategies to keep their social discovery app on top. Like many businesses that need an analytics solution, Tinder initially turned to Hadoop to gain behavioral insights into how users were interacting with the Tinder app. However, Tinder quickly learned that it was very difficult to get the behavior-based insights it needed from Hadoop:

  • First, as data volumes grew, performance quickly dropped off. Queries would take hours to run. Insights were not coming fast enough for the business.
  • Second, Hadoop requires a significant investment in people and machines to implement and maintain. This requirement was a cost and focus that did not align with Tinder’s business strategy.
  • Third, Hadoop and SQL do not provide any native support for behavior-based queries making analysis difficult and lengthy to code – simple questions about how people used Tinder required time from engineers.
  • Fourth, Hadoop and the tools used to query the data are often very difficult to use and required training and specialized knowledge. Departments, like product and marketing, were limited in benefiting from the insights found in the data without significant support.

After using Hadoop, Tinder quickly decided to take a fresh approach to their data strategy. Dan Gould, Vice President of Technology, found Interana to be integrated, specialized for behavioral analytics on event data, and easy for all Tinder employees to use – while still delivering the speed and scale needed to process queries on its massive volumes of data.

Taking a New Approach to Behavioral Analytics

We considered a range of popular analytics options, but none could offer us the speed, scalability and flexibility of Interana and still be truly self-service for both general business users and data analysts. My whole goal with an analytics system is to turn data analysis from a magical temple where only a few people can answer questions into something where everyone throughout the company can quickly and easily answer questions themselves. It’s all about improving the Tinder app, and we need very rapid behavioral analytics to do that. The two big wins of Interana are that’s it’s so fast, and the user interface is surprisingly simple given how much you can do.” – Dan Gould, Vice President Technology, Tinder.

When Tinder found Interana, it was a “Super Like!” Interana was the fresh approach that Tinder wanted for the next evolution of analytics. Tinder immediately saw the benefits:

  • First, Interana scales with Tinder’s rapid growth. Tinder’s cluster already stores over a trillion rows of data while keeping query latency to just seconds. As Tinder continues to be successful, it will not have to worry about its data architecture slowing down because of the growth of data. Additionally, with the ability to keep all its raw data, Tinder can drill deeper into the data for richer insights without any additional ETL aggregation or indexes.
  • Second, Interana is a full-stack solution. By combining the database, analytics layer, and visual frontend, everything works together seamlessly and does not require an army of people to maintain or use. Tinder quickly implemented the solution with its current team, eliminating the need for additional investments in people dedicated to maintenance. This is a huge cost savings for Tinder and keeps the focus on developing Tinder, not data architecture.
  • Third, Interana’s purpose-built behavior-based tools enable the types of analysis that matter to Tinder: conversion, engagement, retention, root-cause analysis and more out-of-the-box. Visual and interactive tools like cohorts, metrics, sessions, funnels, and retention allow Tinder to focus on the types of questions they want to ask – not waste time figuring out how to ask them in complicated query languages.
  • Lastly, the visual interface is simple and interactive, used by multiple lines of business, enabling Tinder to use data more effectively in its daily workflow.

For Gould and the team at Tinder, Interana is at the foundation of its analytics infrastructure, as data is an essential part of the decision-making process. For example, Retention Analysis is critical to how Tinder develops strategies to improve the app. Data analytics is utilized to determine if a feature needs to be promoted or dismantled. The answers and insights lie in the data, and Interana provides access to that information.

With Interana, Tinder can understand how and why key business metrics like engagement, churn, and conversion change over time in seconds. Interana gives Tinder a solution to access our mountains of event data in seconds with a user interface that is surprisingly easy given how much you can do with it,” said Dan Gould, Vice President Technology.

Departments other than the data team at Tinder, use Interana like marketing and engineering. Kyle Miller, Marketing Manager at Tinder, says, “Interana is intuitive with an intelligently designed interface. Before Interana, I would never consider myself a data person, but now I feel like I have the ability to accomplish all of my data-driven tasks.” Miller uses Interana to measure the performance of marketing campaigns, which has traditionally been difficult to do with other solutions. With Interana’s comprehensive and easy to use behavioral analytics tools, along with the scale and speed to analyze all his data, Miller can gain insights previously impossible in record time.

In September of 2015, Tinder launched Super Like, which allows users to signal a new degree of interest in someone by swiping up on their profile. Launching this feature and tracking its use includes a hearty marketing campaign and much-needed event data analytics on the usage of the feature. Interana will be the sole analytics tool used to measure the impact of this campaign and, on the product side, track the usage of this feature to ensure users are getting the best possible experience.

Interana is a daily staple for multiple teams at Tinder. Interactive dashboards and the Retention View allows teams to share their insights and foster greater team collaboration. People start their day sharing new insights and discussing ways to make Tinder even better. Interana’s ease of use is empowering those who never had access to data with insights they never knew existed. These insights, metrics, and other findings are at the foundation of many of the key decision being made today. Users are sharing their knowledge through interactive dashboards and fostering greater team collaboration. From this, Tinder is experiencing faster time to innovation to stay fresh and relevant allowing it to remain the premier social app on the market today. Tinder and Interana – It’s a Match!

Examples of How Tinder is Using Interana

Matching Tinder to the User

Tinder has been hard at work improving the recommendation algorithm. Using Interana, Tinder found that behaviors are different from different genders, orientations and ages in different locations. Through these insights, Tinder has improved the algorithm for different groups of users, making small tweaks and watching as matches increased or decreased.

Better Customer Experience

Ensuring that Tinder delivers the best experience involves more than just adding snappy features — the app has to work. On one occasion, Tinder received reports from a small group of users that Tinder was not working on the 4G network but was working on Wifi. Tinder diagnosed the situation through root-cause analysis using Interana. Tinder found that users were in the same region and shared the same carrier. It turned out the carrier had faulty routing. Tinder worked with the carrier to fix the routing issue and restored 4G service to its users.

A Happy Community of Tinder Users

Sometimes, it is not the app’s operability that is in question, but its intended use. Tinder wants to provide a positive experience for all its users. Interana proved to be successful in achieving this desire when Gould’s team found that a small percentage of users were swiping right on every profile, potentially decreasing the value of matches. After Tinder had analyzed the data to understand what was happening, the company introduced a limit in right swipes each user could perform each day. Since implementing the change, the quality of matches has increased dramatically.

Measuring the Matches Tinder Makes with Other Social Networks

Tinder, leveraging its strong position within the social discovery app community, integrated Instagram into the Tinder app. The Tinder app links Instagram photos to Tinder profiles making it easy to create a new profile on Tinder while allowing users to share more parts of their lives with each other. The marketing teams at Tinder used Interana to measure the success of the partnership. Using Interana, the team received up-to-date information on increased usage and new connections made over time, proving that the alignment between the two social discover apps is a great match.

Segmenting Users into Cohorts for More Advanced Behavioral Analytics

Tinder’s user groups range in demographics from age, location and gender. To understand the behaviors of specific user groups, Tinder creates cohorts of users with Interana and then explores those cohorts to glean comprehensive insights into user behavior. For example, college students inspired the Most Swiped Right Schools list. This list contains names of colleges with high numbers of students who are active Tinder users. When Time published this list, Tinder was able to see where new registrations originated following its publication. Tinder shares information through a dedicated dashboard it created, which is used by multiple teams.

 

Sign up for the free insideBIGDATA newsletter.

Leave a Comment

*

Resource Links: