HotMap

A daily treemap of view-growth momentum across the top-500 performers on Pornhub.

🔗 Live site: hotmap.cam · 📦 Raw data: hotmap.cam/data.json

---

Pages

URL	What
/ (also /rising)	Treemap, default view (Rising Stars × Female × 1d)
/gems	Hidden Gems tier (ranks 251–500) pre-selected
/celebs	Celebrities tier (top 50 by views) pre-selected
/stats	Single-page summary: hero numbers + biggest mover + leaderboards (female-focused)
/charts	A–Z performer index with search + gender filter
/p/<slug>	Per-performer page with sparkline + 1d/7d/30d growth + share buttons
/r/<slug>	CF Worker outbound redirect (click tracking, future affiliate slot)
/data.json	Full snapshot dataset, CC0
/sitemap.xml	All 874+ URLs for search engines

What the treemap shows

Three tiers of performers, ranked by view-growth momentum rather than raw popularity:

Mode	Cohort	Best for
Rising Stars	Ranks 51–250 (middle-tier)	Spotting next-cohort performers gaining traction
Hidden Gems	Ranks 251–500 (smaller accounts)	Discovery of niche / early performers
Celebrities	Top 50 by total views	Tracking the established names

Each tier is sliced by gender (All / Female / Male) and window (1d / 7d / 30d). The treemap colors performers by percentile rank within the cohort — bright green = running ahead of the pack, red = falling behind. Tile size encodes absolute views gained in the window.

A "Top Performer of the Day" card surfaces the strongest mover for the current cohort. Clicking any tile opens the performer's profile (via tracked /r/<slug> redirect).

How it works

1. Daily cron (17 4 * * * UTC) — GitHub Actions runs run.py.
2. Scrape — scraper.py paginates the top-500 list per gender via curl-cffi (Chrome TLS impersonation to pass Cloudflare), fetches each profile for exact view counts and avatar URLs, with retry + auto-backfill for transient failures.
3. Store — db.py writes snapshots to SQLite (data.db, committed to the repo for full history).
4. Render — heatmap.py computes per-window growth, builds 27 precomputed Plotly treemaps (3 modes × 3 genders × 3 windows), the stats and charts pages, and a per-performer page for every slug ever seen.
5. Deploy — workflow commits artifacts back to main. Cloudflare Pages auto-deploys on push; the Worker handles /r/<slug> outbound redirects.

Everything runs on free tiers — GitHub Actions (public repo = unlimited minutes), Cloudflare Pages + Worker (static + edge logic), Cloudflare's automatic SSL. The only ongoing cost is the domain (~$15/year).

Tech stack

• Python 3.13 — scraper, data, render
• curl-cffi — TLS fingerprint impersonation
• selectolax — fast HTML parsing
• pandas — growth-window math
• plotly — treemap and sparkline rendering
• pytest — 26 tests
• Cloudflare Workers — /r/<slug> outbound redirect with click logging
• GitHub Actions — daily cron + commit/push
• Cloudflare Pages — static hosting + CDN + SSL

Local development

git clone https://github.com/ansvier/ph-heatmap
cd ph-heatmap
python3 -m venv venv
./venv/bin/pip install -r requirements.txt

./venv/bin/pytest -q                # 26 tests, ~1 sec
./venv/bin/python run.py            # full scrape + render (~40 min for top-500)
open public/index.html              # eyeball locally

Environment variables:

• PH_IMPERSONATE — curl-cffi browser fingerprint. Default chrome120. Override (chrome119, chrome116) if PH starts rejecting that profile.

Project layout

Path	Responsibility
run.py	Orchestration: scrape → store → render every page → dump
scraper.py	HTTP + paginated top-list + profile + avatar parsing
db.py	SQLite schema, migrations, queries
heatmap.py	Tier filters + treemap + stats + charts + performer page templates
src/worker.js	Cloudflare Worker — /r/<slug> redirect with click logging
wrangler.jsonc	Worker / Pages configuration
public/	Static site served by Cloudflare Pages
public/avatars/	Cached performer avatars (PH blocks hotlinking)
data.db	SQLite store of all daily snapshots, committed to the repo
.github/workflows/daily-scrape.yml	Cron + commit/push automation
tests/	Pytest suite covering db, scraper, render, and growth math

Troubleshooting

• 403 on scrape: Cloudflare/PH is rejecting the GH runner's TLS fingerprint. Set a repo variable PH_IMPERSONATE=chrome119 (or chrome116) in Settings → Secrets and variables → Actions → Variables.
• Could not find 'Video Views': PH changed profile markup. Inspect a profile in browser, update _extract_video_views in scraper.py.
• Avatars missing on a tile: PH layouts differ; _extract_photo_url falls back through #getAvatar and .topProfileHeader img. The next daily run auto-backfills NULL photo_url rows.
• Workflow didn't fire on schedule: GH cron can be delayed during peak hours. We use :17 minute to avoid the busy round-hour slot. Manual trigger via Actions tab if needed.

Data

Snapshot rows are published as JSON at /data.json. Each row:

{
  "snapshot_date": "2026-05-28",
  "slug": "lana-rhoades",
  "name": "Lana Rhoades",
  "total_views": 2123456789,
  "rank": 11,
  "gender": "female"
}

Free to use under CC0. Attribution appreciated but not required. If you build something interesting, let me know.

Disclaimer

HotMap is an independent analytics project. Data is collected from publicly visible Pornhub profile pages via polite (1.5s jittered) scraping. No video content is hosted here. No affiliation with Pornhub or any performer.