Become one of our success stories
Retaining users during major redesign
MySpace was going through one of its major redesigns and we helped parsing the data to observe how users were adopting the changes. In addition to data analysis, we built a ML model to predict churn in parts of the onboarding funnel. The results helped guiding new features and increased retention throughout.
Social network analysis in the metaverse
Flam is the first social metaverse, using Augmented Reality to create unique experiences and interactions between users. We help Flam to understand how users of their first social metaverse are sharing the experience, providing key information for the company to continue improving the product. A key method is based on tracing how users invite other users to the platform and applying social network analysis to find influencers and optimize virality.
Machine Learning for growth
Stepfunction is a leading SaaS company for growth. We conducted R&D by building and evaluating alternative state-of-the-art ML models for their engine, guiding new developments of the product.
Improving the product through analytics
Makers of the popular DoMyShoot app that allows for easy creation of product photos for marketplaces. We helped to specify events to be collected throughout the app to feed an analytics platform. By defining key metrics to track and making them prominently available, the team was able to quickly iterate on features and improve growth, usage and retention.
Reducing friction on bot interactions
Disco (former Growbot) helps companies to build their culture with the use of Slack and MS Teams bots. By using social network analysis, we were able to identify key users that were driving product's adoption in different companies. This information was then used to improve the onboarding process and several features in the product.
Content discovery for kids
Intuary is the creator of Farfaria, an e-reading app for children. We provided comprehensive analysis on how their users were engaging with the app, evaluating their mission of improving reading abilities and progression in children. We also helped developing and evaluating advanced recommendation engines to improve how users discover new content.
User Behavior Analysis
Orkut was an early social network site launched by Google in 2004. While it failed to get traction in the US, it gained an enormous following in Brazil, reaching 30M users in the country by 2008. This number is dwarfed by today's social network sites, but at the time this made Orkut one of the largest social platforms in the world. But why Brazil and, more importantly, how was a mystery. And knowing how could allow Google to expand to other countries.
On Google's request, we conducted a large investigation to uncover this mystery. By using a mixed methods study, where data analysis was complemented by talking to hundreds of users all over the country, we traced the origins and success of the platform to a core and active tech community that was blooming in Brazil and used the platform to connect to each other and to friends and relatives in different states. This core community was so active that essentially solved the cold-start problem for all other users, creating a vibrant platform that is considered to be, to this day, one of the cornerstones of digital literacy in the country.
The findings helped filter countries where similar communities existed and where Orkut could easily expand into. India was initially chosen and a campaign targeting the same type of user responsible for quick-starting the platform in Brazil was initiated. This resulted in over 5M new users joining the platform in less than 3 months, eventually reaching 20M users, making India the second largest country in Orkut after Brazil.
Lolapps rode the app mania that took over Facebook back in 2010 like no other company. It started by making extremely successful quiz apps and quickly expanded into making Facebook games.
These game generated a lot of data and the company needed to find ways to use this data to reduce CAC, increase growth and improve user experience. As part of a team hired to crack the case, we were able to uncover whales, users that spent disproportionally on games. By targeting those users, average LTV rose 10 fold.
By tracing how much users spent in what parts of the game, and how they progressed through them, we were able to propose improved game mechanics that made them more fun for players but also more profitable in the long run. Some results were so interesting that we got a paper published on them.
