Astrolink · eSapiens Tecnologia S.A. · 2011–2018
End-to-end redesign of Brazil's leading astrology social network — from discovery research to app launch.
Product Designer · Product Manager
Context
I joined in November 2016. At that point, mobile already accounted for 42% of platform traffic — but the product had no mobile experience whatsoever. The codebase dated back to 2012.
Year one was about stabilizing what existed: improving flows, fixing friction, and building the foundation for conversion. By the end of it, daily signups had doubled and MRR had reached R$100k/month.
In October 2017, we began structuring the full redesign. Feedback from users was overwhelmingly negative — everyone found it difficult to navigate on mobile. That confirmed what the data already showed.
Then we built the app.
Research
The discovery phase started with a single question written on a whiteboard:
Ensure that people seeking personal guidance perceive recurring value.
Before speaking to a single user, we mapped what success and failure would look like. Success: loyal users, growing revenue, product-market fit, lower churn. Failure: users not converting, not returning, solutions misaligned with real expectations.
We ran quantitative research inside the existing platform using Google Analytics, Hotjar heatmaps, and behavioral data analyzed by our BI team. We combined this with qualitative interviews with users recruited nearby, and usability test sessions with both existing users and newcomers. The sessions were recorded and observed in real time.
In total: 25+ people across all research methods.
Primarily women, 16 to 35. Two profiles emerged: the Apprentice — younger, discovering astrology for the first time, seeking personal guidance. The Expert — experienced, wanting deeper tools and advanced chart reading.
What we found:
Trust barrier
New users hesitated. No clear signal the platform was worth their time — or their birth data.
Onboarding friction
Generating a birth chart requires date, time, and place of birth. Non-negotiable. But the way we asked caused significant drop-off. Users didn't understand why it was needed or how to find it.
Content density
The birth chart is extraordinarily rich. Showing everything at once left users overwhelmed rather than enlightened.
No recurring hook
Users returned when something felt personally relevant. But there was no consistent mechanism to bring them back.
These four insights became the brief.
Approach
The app wasn't a mobile version of the website. It was a rethink of how astrology content could be delivered to a generation that consumes everything on their phone.
Onboarding redesign. Rebuilt from scratch. Each step became a focused micro-interaction, with contextual guidance explaining why birth data was needed and helping users find information they might not have memorized.
Progressive information architecture. Instead of the full birth chart at once, a layered experience. Essential placements first. Deeper content revealed progressively as users explored.
Recurring value surfaces. Daily and weekly content — personalized horoscopes, planetary transits, thought of the day — to give users a consistent reason to return.
Navigation structure. One system to accommodate all astrological tools: birth charts, synastry, tarot, articles, messaging. Designed for a thumb, not a cursor. The hamburger menu was discontinued — a decision backed by frequency-of-use data showing most features were buried without hierarchy or structure.
Prototype and validation cycle. We started with low-fidelity prototypes tested with internal users. Once hypotheses were partially validated, we moved to high-fidelity prototypes and ran multiple rounds of testing with both veteran users and first-time users. The redesigned platform launched first to a restricted beta group, allowing us to analyze KPIs and gather feedback before a gradual rollout to the full user base.
Research: 3 months · Design & development: 3 months (concurrent) · Total: 4–5 months
Impact
The core structure launched in 2018 is still in use today.
People
This project happened because of people who showed up every day and cared about getting it right.
Wladimir built Astrolink from scratch out of pure passion — and trusted us to take it further. Rafael Parigi, Carol, and Guaru brought the product to people in ways design alone never could. Diego, Ícaro, Mauê, Jonas, André, and Washington turned every decision into something real and working.
Thank you.
Learnings
Research changes what you build, not just how you build it. The onboarding problem wasn't obvious until we watched real users struggle with it live.
Constraints produce clarity. Mobile-first forced decisions that a flexible approach would have delayed indefinitely.
The best designs don't need to be replaced. When your structure becomes the foundation others build on for years, that's craft.
Want to talk about this project?
Get in touch →