video analysis+1.5M views
1.5M
views from dating-app data stories
Creators & YouTubers
Content that actually resonates
One creator turned our dataset into two videos on dating-app trends. Nearly 1.5M combined views. The data is the hook.
Trusted by researchersUniversity of Chicago · APA
The dating data that's almost impossible to get.
Anonymized, consent-based behavior from 7,000+ real dating-app users: swipes, matches, and messages. Ready to analyze, publish, and cite. Not a survey.
7,000+profiles analyzed
965k+YouTube views from our data
50+research citations
research-2026-02-23.jsonl
profile · line 1 of 7,214JSONL
{ "profile": { "gender": "MALE", "ageAtUpload": 27, "country": "NO", // Norway "educationLevel": "BACHELORS", "userInterests": ["Travel", "Climbing"], "instagramConnected": true }, "meta": { "swipeLikesTotal": 8432, "matchesTotal": 386, "matchRate": 0.046, "messagesSentTotal": 2914 }, "usage": [{ "dateStamp": "2025-07-12", "appOpens": 14, "swipeLikes": 38, "swipePasses": 91, "matches": 3, "messagesSent": 22, // +11 more fields }, /* +411 daily records */ ], "matches": [/* 386 conversations */] }
Profiles
7,214
Daily records
2.4M
Messages
1.1M
Used by researchers, journalists & creators worldwide
University of ChicagoYouTube creatorsData journalistsData scientists
Why it's different
Real behavior, not self-reports
Dating apps don't share user data, and surveys can't capture what people actually do. SwipeStats is built from users' own GDPR exports: the ground truth of how people swipe, match, and talk.
Observed, not claimed
Actual swipes, matches, and messages with timestamps, not what people say they do on a questionnaire.
Anonymized & consent-based
No names or contact info. PII in bios and messages is LLM-redacted. Every profile was voluntarily uploaded by its owner.
Rich, computed data model
Five linked objects per profile, plus pre-computed metrics like match rate, response time, and ghosting. Ready to analyze.
Continuously refreshed
New profiles arrive every week as users upload their exports. Fresh tiers ship the most recent activity available.
Typical survey data
- Self-reported and memory-biased
- Small N, expensive to field
- No longitudinal behavior over time
- Can't see real messages
SwipeStats datasets
- Logged behavior, straight from the source
- 7,000+ profiles, growing weekly
- Day-by-day usage spanning years
- Anonymized message-level text
The data model
Five linked objects in every profile
Each line of the JSONL file is one user. Raw profile fields, geography, pre-computed stats, daily activity, and every conversation. All keyed together.
profile28 fields
Demographics, bio, interests, education, search filters.
user6 fields
Resolved geography, languages, and timezone.
meta24 metrics
Pre-computed match rate, response time, ghosting & more.
usage[]15 / day
One record per active day: swipes, matches, messages.
matches[]+ messages
Every conversation with message-level, redacted text.
Sample rows: meta5 of 7,214
| gender | age | likeRate | matchRate | ghosted |
|---|---|---|---|---|
| MALE | 27 | 0.41 | 0.046 | 165 |
| FEMALE | 24 | 0.18 | 0.312 | 88 |
| MALE | 31 | 0.63 | 0.029 | 241 |
| FEMALE | 29 | 0.22 | 0.287 | 112 |
| OTHER | 26 | 0.34 | 0.094 | 73 |
Everything is documented. The full data dictionary defines all 70+ variables, their types, nullability, and how each metric is computed, with notes on PII redaction and data quality.
Format
JSONL
Python · R · pandas ready
Per profile
70+
documented variables
See it before you buy
A glimpse of what's inside
One real, anonymized profile from the demo dataset, rendered exactly as it ships. Walk a single row end to end before you buy.
The exact dashboard a buyer gets. Click through one real row.
Use cases
From viral videos to peer-reviewed research
manuscript figure
n = 7,214 profiles
effect size
Researchers & academics
Publish credible, real-world research
University of Chicago and others have used SwipeStats for studies on modern dating. Thousands of profiles, statistical power.
data story
The patterns surveys miss
swipesmatchesmessages
Data journalists & writers
Data-driven stories that get read
Writers have analyzed hundreds of profiles, including messages, to surface patterns no survey could reveal.
notebook.ipynb
df.groupby("age")
.agg("matchRate")
# portfolio-ready dataset
run 12
Data scientists & hobbyists
Build a portfolio on unique data
Practice analysis and visualization on a domain everyone relates to. A standout portfolio project with a real, rich dataset.
Pricing
Choose your dataset
For a blog, a paper, or plain curiosity, a SwipeStats dataset gets you on the right track. Start free.
Starter Pack
Test your hypothesis with real data.
$15
- 10 profiles
- Email support
- Personal use
- Great for small projects
StandardBest value
The go-to for creators and researchers.
$50
- 1,000 profiles
- Commercial use
- Publication rights
- $0.05 / profile
FreshMost popular
The most recent data available.
$150
- 1,000 newest profiles
- Priority support
- Commercial + publication
- Latest dating trends
Premium3,000 profiles
Serious research with statistical significance.
$300
- 3,000 newest profiles
- Priority support
- Deep market analysis
- Commercial + publication
Academic LicenseInstitutions
For universities and institutional research.
From $1,500
- 5,000+ profiles
- Custom data requests
- Student distribution rights
- Monthly ongoing support
Compare
Every plan, side by side
| Feature | Free sample | Starter | Standard | Fresh | Premium | Academic |
|---|---|---|---|---|---|---|
| Price | $0 | $15 | $50 | $150 | $300 | $1,500+ |
| Profiles | 1 | 10 | 1,000 | 1,000 | 3,000 | 5,000+ |
| Price / profile | $0 | $1.50 | $0.05 | $0.15 | $0.10 | $0.30 |
| Data recency | Sample | Mixed | Mixed | Newest | Newest | By request |
| Email support | — | ✓ | ✓ | Priority | Priority | Priority |
| Commercial use | — | — | ✓ | ✓ | ✓ | ✓ |
| Publication rights | — | — | ✓ | ✓ | ✓ | ✓ |
| Student distribution | — | — | — | — | — | ✓ |
How it works
From checkout to analysis in minutes
1
Choose your dataset
Pick the package that fits your project and budget. Not sure? Start with the free sample profile.
2
Instant download
Most datasets download immediately with a license key. Academic licenses are processed within 24 hours.
3
Start analyzing
Load the JSONL into Python, R, or pandas. The full data dictionary ships with every download.
FAQ
Questions, answered
Is this data ethical and legal?
Yes. Every profile is voluntarily uploaded by its owner via their official GDPR data export, and fully anonymized. No names or contact details are included, and PII in bios and messages is LLM-redacted. We comply with applicable data-protection regulations.
What format is the data in?
JSONL: one JSON object per line, one line per profile. It imports cleanly into Python, R, pandas, or any tool you prefer. A full data dictionary defining every field ships with each download.
Can I use this for commercial projects?
Yes. The Standard tier and above include full commercial-use rights. Blog posts, YouTube videos, paid research, or any commercial purpose are covered. Check your tier for specifics.
Can I publish research using this data?
Absolutely. Standard tier and above include publication rights; we just ask that you cite SwipeStats as your data source. Researchers and journalists already have. See the documentation for the recommended citation format.
How recent is the data?
Starter and Standard mix timeframes. Fresh and Premium ship the most recent profiles available. Academic licenses can request specific periods. New data is added continuously as users upload their exports.
How do I cite this in research?
We provide a standard citation with every download. Generally: "SwipeStats.io Dating App Dataset, [Year], [Number of Profiles]". The documentation page has APA and BibTeX formats ready to copy.
Get started
Start exploring real dating data today
Join the researchers, creators, and data enthusiasts uncovering how modern dating actually works.