Reserhub turned its product team into a data-driven organization with Mixpanel, usage-frequency segmentation, and ongoing learning rituals supported by Bildung.
The context
Reserhub is an intelligent platform for bus ticket management and sales that powers the direct digital channels of transportation companies across Latin America. With over eight years of experience, its product and design team had been working with Mixpanel, FullStory, and Hotjar to understand user behavior. The real challenge, however, wasn’t the tools — it was getting every team comfortable exploring data and weaving it into their day-to-day work.
From the very beginning we understood the value of knowing how users interact with our products. It was a maturity journey that took time, but today data is part of the day-to-day. — Jair Pérez, Head of Product Design at Reserhub
The challenge: from intuition to data
Reserhub already had analytics tools in place, but the real challenge was cultural: helping people across functions get familiar with events and reports, lose their fear of exploration, and translate that data into product decisions. Mixpanel in particular could feel overwhelming for new joiners.
Mixpanel can feel overwhelming at first. One of the main challenges was helping new team members get familiar with our events and reports, and overcome the fear of exploring. — Jair Pérez
The solution with Bildung: segmentation and Mixpanel adoption
Bildung’s engagement with Reserhub focused on a clear goal: help the team get the most out of Mixpanel and extend its use beyond the product team. As a Mixpanel implementation partner, Bildung worked alongside Reserhub to define new analytics practices, refine event taxonomy, and — above all — build a usage-frequency-based user segmentation that reshaped how the team understands real user behavior.
The segmentation was structured into three groups: Power users (90th percentile and above — those who use or buy the most), Core users (median or mid-range frequency), and Casual users (25th percentile or below). This structure didn’t just help describe cohorts; it unlocked sharper business questions: Is there a correlation between frequency and retention? Which patterns separate power users from the rest? Which features push casual users into the core segment?
Experimentation culture: GrowthBook + Mixpanel
A/B testing became one of the pillars of Reserhub’s new data culture. Whenever a new feature ships, the team uses GrowthBook to split audiences and Mixpanel to measure real behavior and impact. Typical rollouts start by releasing the new flow to 20% of traffic and, once stable, open up to a 50/50 split to compare versions.
We try to ship every new feature with an A/B test. We rely on GrowthBook to split versions and Mixpanel to track behavior. It’s not just about shipping — it’s about learning which version actually improves the experience. — Jair Pérez
This approach helped optimize conversion in key flows, including the schedule and seat selection experience. Today, most releases at Reserhub include some form of experiment, and reviewing results is a natural part of closing out each release.
Data rituals and informed decisions
To sustain adoption beyond the initial implementation, the team built recurring rituals: monthly insight reviews, biweekly exploration sessions, and cross-team reviews between design, product, and customer success. Combined with a clear taxonomy and ongoing training, these rituals turned data into a shared language across the organization.
The data culture at Reserhub is growing. Informed decision-making is more and more valued, and there’s genuine interest in having accessible, reliable data. — Jair Pérez
Key takeaways
Reserhub’s story leaves three lessons for teams just getting started: document events and properties so everyone knows what is measured and why; foster curiosity by asking what you want to learn before each release; and invest in continuous training to build analytical judgment, not just technical skill. A data culture isn’t bought — it’s built, with time, curiosity, and practice.
