Today, about one-third of Americans have used a dating app or site, and 12% have either been in a committed relationship or gotten married to someone they met through online dating, according to a recent Pew Research report. Meeting the right person may seem like magic, but if you’re using a dating app or website, meeting the right person is a calculated process. Online dating has always been a data-driven, scientific, and effective way of connecting people who share common goals and interests.
There are plenty of online dating apps that have sprung up over the years, catering to just about every interest, community, and affiliation. OkCupid has been around since the beginning, and today, OkCupid’s use of business intelligence (BI) and product analytics tools are behind the platform’s success.
Driven by Data, Powered by the Heart
Data is core to the mission here at OkCupid. Our data obsession is why OkCupid makes more than 4 million connections every week, over 200 million a year, 5 million introductions a day, and gets more mentions in the New York Times wedding section than any other dating app.
I’ve been with OkCupid for three years and I manage our data science team, which handles platform analytics. It’s exciting to see meaningful human connections develop, but it’s rare to open a dating app and immediately find love. Users have to stick around for a while so the app can learn their likes, dislikes, deal-breakers, and other information to help locate a compatible match.
One of OkCupid’s key differentiators is the use of questions to create a match score that determines one person’s compatibility with someone else. The more questions we ask, the more information we receive, and the better we can pair users with someone else. To do this, however, we have to understand the mountains of data we obtain.
Creating the Perfect Data Stack
The focus of the data analytics team is to understand how the OkCupid platform functions and what we can do to improve it. Our work ranges from traditional business intelligence (BI) reporting to algorithm development and optimization with a macro focus on user experience (UX) and product optimization.
Our customer data stack at OkCupid consists of mParticle, Looker, and product intelligence (PI) platform Amplitude. mParticle collects and stores our customer event data, which we send to Looker for general business reporting, and to Amplitude for deeper analysis on user behavior and our customer experience.
When my team first started using Amplitude, we had this conception that it was mostly for event tracking and segmentation. Eventually, we learned that we could use it to measure engagement, to identify user cohorts, to analyze different user journeys, and to find leading indicators of conversion and retention. Amplitude is explicitly designed for this type of analysis, which meant we could access meaningful insights that much faster.
BI and Amplitude: Better Together
Building the most engaging and enjoyable product possible requires a lot of A/B testing and data analysis to determine what aspects of our product customers like, and find opportunities to boost engagement with them. Whether it’s a high-intent user looking for a long-term committed relationship, or an occasional user looking for something more casual, we have to understand who those different users are, the different ways that they engage with the platform, and the behaviors and motivations that cause them to stick with the platform or drop off over time.
Traditional BI tools like Looker, Tableau, or Power BI, can perform this analysis, but they require us to spend time building out data models to answer our product questions. They also have their limitations when it comes to the depth of insights we can glean from the data we have.
Improving the #UserJourney requires more than high-level insights. You need detailed, coherent information about user sessions. Click To Tweet
With Amplitude, we can make sense of unstructured data and start to understand our different users and their journeys in our product. From there, we can build out more structured reporting, identify the product experiences that customers find most valuable, and build more of them into OkCupid.
For instance, Amplitude allows us to identify and understand the various behaviors that indicate users will spend a long time in the app. And for those users who log on and then quickly leave the app, Amplitude provides us with user pathways that we can analyze to see what happens most often before a user ends their session. As a result, we can figure out what aspects of OkCupid we should change—or remove altogether.
A traditional BI tool like Looker can access all of the information in our data warehouse, and run traditional aggregations and pivots very easily. But Amplitude shines when handling time-series events and anything that is not well-structured.
To give a concrete example, it’s easy to use a BI tool to answer, “How many likes has a user made over time?” Where Amplitude provides additional value is in understanding what led them through that journey to those likes. Did they come in through a notification or by navigating through different parts of the app? Where did they go from there and what was their typical engagement pattern with various features? So instead of just knowing that a user liked 20 people today, we can begin to form a story about that user’s experience and preferences. Maybe they liked 20 people today, and spent a lot of time sending messages to each of them, which is different from someone who liked 20 people today, but did so in rapid succession.
The nuances in our customers’ experiences are hard to see when we’re doing aggregates. Looker is built on incumbent data storage systems, so to answer a question like above, you’d need to build a custom report, join together multiple data sets, or even write SQL. When using Amplitude, the differences are easy to see when we have that user journey in front of us.
Better Teamwork and Faster Launches
Our main users of Amplitude are my data science team and our product teams. Both groups ask questions around user journeys and engagement, but they need answers to different types of questions, too. For instance, we have a team dedicated to our onboarding flow, and they care about drop-off points for new users. Another team focuses more on long-term retention, so they care much more about sticky behaviors, those that keep people returning to the site and creating a better chance of success in love.
Amplitude allows us to create and save all our various charts and dashboards, and pollinate them across the organization. This means we don’t have to duplicate efforts; teams share results regularly and make decisions from the same data set. Even though we have a self-serve approach to our data, it’s a truly collaborative process that saves us time, and leads to more informed decisions.
Using tools that combine teamwork and functionality means you can launch new features much faster. Click To Tweet
Amplitude allows us to look at the structured data without spending the additional development time required to build out new views. Whenever we launch a new feature, we simply instrument an event for it in mParticle, and send it to Amplitude with the appropriate user and event properties. Traditionally, ensuring accurate data tracking within the platform would’ve required an analyst writing manual queries in Python or SQL. With Amplitude, we don’t need an analyst’s help. We can see the new events loading in real time, and immediately QA them in Amplitude charts.
The ultimate impact of that powerful combination of teamwork and functionality is that we can immediately understand whether a product bet is working or not, and iterate much faster than before.
Concrete Direction in Times of Change
Widespread lockdowns have shifted dating norms this year. Rather than bemoan the loss of traditional dating protocols, we had a new question to answer: How are people changing their usage patterns on our platform to adapt to a COVID-19 dating world?
Finding love and connection isn’t as difficult when it’s data-driven. Understanding the #OnlineDating #UserJourney is the key to your users’ hearts. Click To Tweet
For starters, we’ve seen that users spend a lot more time digging deeply into conversations. People can’t meet in person as easily as before, so they have to spend more time getting to know someone within the app itself. Using our powerful combination of BI and PI, we can quickly access concrete data on these new patterns. By creating even more opportunities for users to forge deeper virtual connections, we can fill some of the void that physical distancing has created for a lot of people.
Finding love through data analytics might not seem romantic, but we know that it works. OkCupid’s mission has always been to bring love to the world, and with the right data applied in the right way, we help people to do just that.
