# Decoding the science of attraction

We partner with top researchers to study attraction, compatibility, and what makes relationships last.

We provide match data, outcome metrics, and in-product experiments. Our engineers build custom surveys, psychometric tools, and longitudinal tracking. Everything is privacy-first and IRB-friendly.

[Submit research proposal](#)

## Meet the experts advancing Keeper’s research

Geoffrey Miller, PhD

Expert in human mate choice, signaling theory, and evolutionary psychology

Bestselling author of _The Mating Mind, Spent,_ and _Mate_

Shapes Keeper’s scientific framework for attraction and partner matching

[Learn about Geoffrey](https://en.wikipedia.org/wiki/Geoffrey_Miller_\(psychologist\))

Michal Kosinski, PhD

Pioneer in computational psychometrics and digital behavior prediction

Proved that digital footprints can accurately predict personality traits, preferences, and behaviors at scale

Guides Keeper’s psychometric modeling and data-driven insights

[Learn about Michal](https://en.wikipedia.org/wiki/Michal_Kosinski)

Naman Gupta, PhD Student

PhD researcher, Stanford Management Science & Engineering

Specializes in dating market analysis, optimization, and algorithm design for two-sided matching

Leads internal studies on relationship formation predictors and optimizes Keeper’s matching algorithms

[Learn about Naman](https://msande.stanford.edu/people/naman-gupta)

One of the richest relationship datasets anywhere

1.6M users and growing

Combining structured data with deep, long-form responses

Age

Height

Politics

Religion

Gender

Big  5  traits

additional  personality measures

IQ

lifestyle  habits

Exercise  routines

Dietary  preferences

Pet  ownership

Attachment  Styles

and  much  more

Deeply committed to privacy

We use privacy safeguards at every stage

Depending on the study, research may use aggregate, deidentified, anonymized, or pseudonymized data. We minimize the information used, remove direct identifiers, restrict access, and require research/model development partners to protect the data, use it only for approved purposes, and avoid reidentification.  
  
Sensitive traits are used or shared for research/model-development only where allowed by law and with consent where required. Published research and public insights are presented in aggregate or deidentified form so they do not identify individual users.

[Submit research proposal](#)