Data Ethics
Relationship data is sacred.
We treat it that way.
The data people share about their relationships is some of the most personal information that exists. Our entire research model is built on the premise that this data deserves extraordinary care.
Three-tier consent model
Not all data is treated the same. Our consent model has three distinct levels, each with increasing transparency and participant control.
Tier 1
Product Operation
Required to use the platform
- Data used solely to deliver the product experience
- Account information, preferences, and feature usage
- Never shared externally in any form
- Standard data handling per privacy policy
Tier 2
Anonymized Research
Opt-in at signup, toggle off anytime
- Behavioral patterns aggregated across thousands of users
- Fully anonymized at the point of collection
- No individual identification possible
- Powers our published research and quarterly reports
Tier 3
Identified Research Participation
Separate opt-in, compensated
- Voluntary participation in specific research studies
- Participants are informed of exact study purpose
- Compensation provided for time and contribution
- Full withdrawal rights at any point, data deleted on request
Our data principles
These aren't aspirations. They're constraints we build into every system, every pipeline, every research protocol.
Consent is architecture, not afterthought
Every data point in our system carries a consent tag from the moment it's created. We don't collect first and ask permission later. Consent level determines what we can do with each piece of data, enforced at the infrastructure level.
Anonymization at the point of collection
Tier 2 research data is stripped of identifying information before it enters our research pipeline. We don't anonymize later. We anonymize first. The research team never sees individual-level data.
No individual predictions for external use
We study patterns across populations, not individuals. Our research identifies what couples in general do differently. We never build models that predict outcomes for a specific person or couple for any external partner.
Withdrawal is immediate and complete
If a participant withdraws consent, their data is removed from all active datasets. For Tier 2 aggregate data already published, individual contributions cannot be separated. But no new analysis will include their data.
We publish limitations, not just findings
Every research report includes sample size, methodology, confidence intervals, and known limitations. If the data doesn't support a strong conclusion, we say that. Overstating findings would undermine the trust our entire model depends on.
GDPR strictness as the global baseline
Regardless of where our users are located, we apply GDPR-level data protection standards to all participants. Right to access, right to deletion, right to portability. Not because we have to. Because it's the right architecture.
Our commitment
We understand that people will only share their most personal data with organizations they trust completely. That trust is earned through behavior, not promises.
Every decision we make about data handling is guided by a simple question: would we be comfortable if every participant could see exactly how their data is being used? If the answer isn't an unqualified yes, we don't do it.
This page will evolve as our practices evolve. We'll document changes publicly and explain our reasoning. Transparency isn't a feature. It's the foundation.