Score-Watch Alert
Email the moment a tracked entity’s Compassion Benchmark score moves. Delta, band change, headline evidence, and a link to the full assessment — delivered the morning after overnight research flags the change.
What you get
A concrete, per-entity signal — not another weekly newsletter.
Per-entity email alert
One email per score change for the entity you’re watching. No generalist digest. No marketing blasts.
Delta, band change, evidence
Every alert includes the composite delta, whether the entity crossed a band boundary (e.g., Functional → Developing), and the headline evidence behind the change.
Link to full assessment
Click through to the daily briefing with full evidence, dimension-level analysis, and confidence score.
Overnight research, six days a week
The pipeline scans 1,155 entities for evidence every night Monday–Saturday. Your alerts land in your inbox the morning after a change is confirmed.
Independence-preserving
Subscription purchase does not affect the score. Entities never pay for inclusion, score changes, or suppression. See the independence policy.
Cancel anytime
No lock-in. Stop the alert at any point. Refund on the unused portion of the year, subject to the refund policy below.
How it works
Three steps. No configuration. No dashboards.
Pick an entity
Browse any of the seven indexes below and find the entity you want to watch. Any assessed entity is eligible — companies, countries, labs, states, cities.
Subscribe from the entity page
On the entity’s detail page, click Subscribe — $79/yr. The subscription is scoped to that entity. Add more entities by subscribing again.
Receive alerts when scores move
Overnight research runs six nights a week. When the assessment moves an entity you’re watching, an alert is in your inbox by morning.
Eligible entities — every benchmark index
1,155 entities across seven indexes. Every assessed entity has a detail page and can be watched.
Who buys Score-Watch
Teams and individuals who need to track a specific institution continuously, not periodically.
Investors and analysts
Tracking ESG risk signals for one or more portfolio companies. $79/yr per name is a fraction of one analyst hour — and the signal lands before it’s in the news cycle.
Policy and trust-and-safety teams
Watching AI labs, robotics companies, or platform operators for governance signals that affect policy work. Alerts include legislative context when relevant.
NGOs and journalists
Tracking institutions on your beat. Every alert is evidence-linked and citable. Source list included.
Executives at benchmarked institutions
Know when your own score moves, what triggered it, and what evidence is now public — before a stakeholder asks.
Refund policy
We stand behind Score-Watch. If the product doesn't deliver, we refund.
- Within 14 daysFull refund if no alert has been delivered during the subscription period. Request within 14 days of purchase date.
- 15–90 daysPro-rated refund for the unused portion of the year (unused months × $6.58/mo) if requested after the first alert has been delivered, up to 90 days from purchase.
- After 90 daysNo refund after 90 days from purchase. The subscription continues until the annual expiry date.
To request a refund, email alerts@compassionbenchmark.com. Refund requests are processed manually within 5 business days via Gumroad refund.
Independence safeguards
Score-Watch is an observer product. It is structurally incapable of influencing what it tracks.
Two separate technical planes
The assessment pipeline (which produces scores) and the commercial system (which sells Score-Watch) run in separate technical planes with no shared credentials or network paths. The pipeline cannot read subscriber state; the commercial system cannot write scores. See methodology.
Alerts fire from research output only
The alert sender reads exclusively from the overnight research pipeline’s change-proposal files. No commercial event — purchase, cancellation, renewal — can trigger or suppress an alert. Subscriber lists are invisible to the scoring system.
Auditable at the file level
Every score change is sourced to a git-committed change-proposal file with a proposal ID. Alert emails cite the proposal ID. The commercial system has no GitHub write token and cannot modify score files. Independence methodology.
Start with one entity
Most subscribers start with one entity and add more as the need becomes concrete. No annual commitment beyond the one subscription. Cancel anytime.