Kast uses Mixpanel to identify their power users and understand what leads to stickiness

In the unprecedented times brought on by COVID-19, the world has had to quickly adapt to a new normal that often feels anything but. With a collective responsibility to socially distance from friends, family, and colleagues, many have struggled as they attempt to navigate how to stay connected while being apart. Enter Kast, a “virtual living room” that allows community members to participate in everything from family game nights to study dates—just like old times.

Originally conceptualized as a screen-sharing platform for esports teams, the beta phase quickly revealed that people were turning to Kast for so much more. After officially launching in July of 2019, Kast has been used by millions of people seeking to replicate the magic of an in-person interaction in a digital setting. For Co-Founder and CEO, Dr. Mark Ollila, facilitating those meaningful connections is at the heart of everything he does. The feedback has been more validating than ever, with many community members crediting Kast as the only thing saving their social lives during quarantine. And yet, the satisfaction of a job well done is the farthest thing from Mark’s mind; instead, he is more invigorated and inspired than ever in his tireless pursuit to innovate.

Early on, one of Kast’s most notable challenges was that, more often than less, decisions big and small were made on feelings, not facts. As the company’s growth skyrocketed, they knew they needed to scale; unable to discern what aspects of the product were informing the ways in which their community members engaged with it, decisions began to feel like a shot in the dark, and it was as hard to understand and learn from mistakes as it was to identify what they were doing right. When Mika Kunto joined as Kast’s Director of Innovation & Product Technology, he resolved to change things, quickly cementing the phrase “data wins arguments” as the company’s new guiding principle. With the right approach finally in place, all that was left to do was to find the right tools to realize it.

After carefully evaluating and prioritizing the team’s goals and challenges with the company CTO Tim Flack, Mixpanel was selected as the tool that would help the Kast team become more data-informed. Not only was he impressed with Mixpanel’s ability to analyze conversion, engagement, and retention, which he believed were key to making better product decisions faster, but Mixpanel had also received glowing endorsements from several team members who had previously used the tool—the choice felt obvious.

“Mixpanel validates our decision-making,” says Chief Experience Officer, Elina Ollila. When a long-standing hypothesis that Push to Talk, a feature originally conceptualized during the genesis of what was once meant to be a platform for game players to interact, was not relevant for the majority of community who were instead using the product for activities like movie viewing parties, it seemed like the obvious to retire it. Instead, they dug into the data using Mixpanel, and found that the feature was in fact one of their widely used—precisely the opposite of what they had expected to see. Dr. Ollila explains, “I’m so glad we looked at the Mixpanel data first, we dodged a bullet. If we had removed Push to Talk, we would have gotten community outrage. The minimal time the team spent digging into the data was nothing compared to the kind of firefighting down in the community side if things went wrong.”

With an abundantly active community, Dr. Ollila likens Kast’s relationship with them to a marriage, explaining, “there’s a certain amount of mistakes we can make, but if you keep making them over and over again, they’re going to leave.” To prevent that from happening, the team relies heavily on qualitative feedback from community members in addition to Mixpanel data. In the past, the team would receive survey feedback from users who wanted to express their view but weren’t familiar with the product. After using Mixpanel’s cohorts feature, they were able to create a segment of “power users,” defined as community members that have both been active in the last 15 days and entered a party at least 30 times. Now, Kast relies on this specific cohort (in addition to others) to gather their product feedback, trusting that it will be more relevant, valuable, and actionable. “It’s great to build a segment of people that we trust with their qualitative input and feedback on product direction, Mixpanel cohorts help in finding those people who are likely to have the most valuable feedback for each research question.” Dr. Ollila shares.

“Mixpanel also allows our developers to spend time on building the right features,”Dr. Ollila continues. The team was rolling out an array of new emoji reactions that they thought would enhance people’s interactions during virtual parties. After consulting the data in Mixpanel, they quickly realized that while emojis associated with negative reactions weren’t getting any traction among community members, ones associated with positive reactions, like animated hearts, or relevant to the occasions and current events, like tissue paper, were getting tons of engagement. She concludes, “the information we’ve uncovered in Mixpanel has informed us what products and features we should invest in and create in the future.”

Moving quickly is something the Kast team is accustomed to; in less than a year, the product has launched and become an international sensation. Mr. Kunto credits this increased precision and agility to Mixpanel, citing that “Mixpanel data enables me to understand product/features better and helps us to drive better product decisions”. New data and insights would give us a tool to prioritize new product/feature development.” With detailed and immediate access to insights, he can quickly analyze and understand the impact—negative or positive—of the features they ship, which in turn informs what products they will build next. Having used an in-house tool in the past, Kunto feels. “Mixpanel is even better because it would give us capability to instrument assets fast and get quick product feedback. When you don’t have to start from scratch it would speed up new features development and save valuable time” Dr. Ollila shares this sentiment, explaining that “before Mixpanel, the team didn’t have insights into how features were being used or what was most popular. Mixpanel has given insights unlike we’ve ever had before.”

Unified view of customer data

Prior to Mixpanel, Kast was unable to see all the data associated with a community member in a singular, easily accessible place. Although they were able to see certain types of data, like total active users or numbers of parties attended, the data they were getting largely was around traffic, and missing context, making it neither reliable nor actionable.
With Mixpanel’s ability to track user data, events, and properties, Kast can easily access all manners of information—ranging from which party a community member has attended to the country from which they are accessing the platform—all in one place. With a holistic view of their community members, Kast is able to see what people are doing, how long they’re doing it, and how they’re moving through the platform. Most importantly, they can uncover the behavioral correlations between the actions of their most frequent and infrequent groups of community members. These insights empower the team to make more deliberate, informed product decisions in order to drive engagement and retention.

Understanding of stickiness

For Kast, a big benefit of Mixpanel is its ability to help them identify and understand indicators of stickiness. The Retention report is an excellent starting point to start building stickiness indicators and has helped the team build more dedicated insight and metrics about user behavior in the platform. Kunto elaborates, “we’ve started to look heavily at Week 1, Week 4, and Week 12 retention queries in Mixpanel, and since, we’ve realized that community members don’t become committed to our platform overnight but it would take days or even weeks to build relationships.” As a result of this insight, the team developed a “Churn 30” metric, which they define as users who churned 30 days ago that have organically returned divided by the total number of people that churned 30 days ago. With metrics such as this guiding their strategy and product management to focus activities that helped increase retention.

Set up for success with Mixpanel Professional Services

The Kast team worked closely with Mixpanel’s Professional Services in order to create a strong implementation. Today, the Kast team remains impressed with the level of service they receive; for the questions they are unable to answer themselves within the product, they have the peace of mind of knowing they have a dedicated team at Mixpanel helping them at every turn.

Shifted company culture

In the beginning, a lot of decisions were based on a gut reaction now the team depends on data. Being data-driven is now a hallmark of Kast’s company culture. In keeping with this, Mark looks forward to receiving his daily email digest from Mixpanel, tracking activated users, monthly actives, weekly actives over monthly active user percentage, churn 30, weekly retention, and more. Better still, democratizing this information across the entire organization has allowed each member of the team to be an informed actor on their data, servicing Kast’s ultimate goal of delivering superior products and experiences to their users faster than ever before.