Contact

Single inbound channel: bench@aiporngirlfriends.com.

We route messages by subject prefix. Use the prefix that applies; messages without a prefix are read but deprioritised.

[methodology] · methodology questions, weight-vector queries
[dispute] · factual disputes on a published composite or sub-score
[data] · raw-data requests, latency probe logs, historical cycles
[platform] · platform vendors regarding their own listing
[reader] · readers, with suggestions or questions

What we respond to

• Methodology questions, we publish clarifications publicly when the same question recurs.
• Factual disputes, we re-run the affected probe and either correct the published value or document why the original stands. Reply target: 5 business days.
• Raw-data requests, weekly probe dumps are already public at /data/<year>-<month>.json. Custom-cut requests are evaluated case-by-case.
• Reader suggestions, we read everything. We do not promise coverage but suggestions inform shortlists.

What we do not respond to

• PR pitches, however carefully framed.
• Pre-publication outreach from platforms about pending coverage.
• Requests to “review” the methodology privately before publication.
• Score-negotiation attempts of any kind.

For platforms

If your platform has been benchmarked and you believe a sub-score is factually wrong, write with the prefix [dispute], cite the cycle and sub-score, and provide documentary support for the corrected value. Disputes that consist of “we disagree with the score” without documentary support are read and filed; we do not amend scores without evidence.