Skip to main content
effective 2026-08-14

Privacy

There are no accounts, and no data is sold

The data Scrutica collects and who processes it.

Scrutica is a public research site. This page describes its data practices as of 2026-08-14; material changes are recorded in the Record, the site’s log of what changed and when.

What is collected

Usage analytics
Page views and in-page interactions are measured with PostHog, Google Analytics, and Vercel’s Web Analytics and Speed Insights: pages visited, referring site, browser and device class, page-performance timings, and approximate (city-level) location derived from IP address. These measurements show which pages researchers use. The analytics configuration uses no advertising identifiers or cross-site tracking, and this data is not shared for marketing; Google Signals and ads personalization are disabled.
Session replay
PostHog additionally records every session as a reconstruction of the pages you saw and the clicks, scrolls, and console output within them. Anything you type into a form is masked in the replay, and network request bodies are not recorded, so the contents of the subscription and suggestion forms do not appear in it. Replays show where a page confuses people.
Cookies
Three first-party cookies persist between visits: two set for Google Analytics (up to thirteen months) and one for PostHog (up to twelve months). They distinguish returning browsers for the analytics above, and PostHog mirrors its identifier in local storage for the same purpose; nothing else reads them.
Programmatic-access telemetry
Calls to the MCP server and agent fetches of machine-readable surfaces are logged with the tool or path invoked, the client name and version it announces, latency, and status. Identity is represented by a one-way hash or the bot family name, never by a stored raw address tied to a profile.
What you submit
The email subscription form stores the address you enter, used only to send the updates you asked for; write to the contact below to be removed. The data-correction suggestion form stores what you submit for editorial review. Neither is shared.
Transient operational data
Rate limiting keeps per-address request counters for minutes. Infrastructure providers also retain standard hosting logs.

What is not collected

There are no user accounts, so there are no passwords, names, or account profiles. Reading preferences that persist between visits live in your own browser’s local storage and are never transmitted to us. We do not run third-party advertising, we do not embed social-media trackers, and we never sell personal data.

Who processes data on our behalf

Vercel (hosting and delivery, plus its Web Analytics and Speed Insights), Supabase (database), PostHog (analytics and session replay), Google (analytics), Upstash (rate-limit counters), and Mapbox (map tiles). On map pages, your browser fetches tiles directly from Mapbox, which therefore sees the request. Each processor uses data solely to provide its service to us.

Retention and your choices

Analytics and telemetry are held by the processors named above, under the retention each of them is configured for; none of it is written to Scrutica’s own database. The site works in full with an ad blocker or browser tracking protection switched on.

To remove a subscription address or a suggestion you submitted, or to ask anything about this policy: davidgringras@hsph.harvard.edu.

Agents and crawlers

Automated clients are welcome; Scrutica for agents lists the endpoints and the rate limits. Their requests are logged under the same anonymous telemetry described above. The data covers organizations and facilities, not private individuals.