Trust is the product. So we made it checkable.

Every quote is checked against the audio.

Our pipeline transcribes the captured recording and matches every candidate quote back against it. A quote that cannot be matched to the audio never ships. Anything the model produced that the recording does not support is dropped, along with its signal.

You do not have to trust our summary. Every signal carries the clip, and you can hear the executive say it. That is the whole trust model: the clip is the receipt.

Public earnings audio only.

Earnings calls are public events. Companies invite the market, publish the audio, and answer analyst questions on the record. We capture the public webcast, including registration-walled ones, the same way any attendee does, and we work from the same material available to every analyst and journalist.

We do not scrape private communities, record private conversations, or reintroduce licensed transcript archives. The dataset is built forward from our own capture.

No contact data stored. Ever.

We store no emails, no phone numbers, no scraped profiles, no org charts. Names appear in our data only as public call speakers, with their public role, the way a newspaper quotes an executive.

Bring your own enrichment. You pair our why-now signal with the contact data you already license from ZoomInfo, Apollo, or your provider of choice, under your agreement with them. Your compliance team keeps one clean answer: GTM Signals holds no personal contact records.

We never resell transcripts.

What leaves our system is the derived signal: one verbatim quote sentence, a why-now, a relevance score, and our own audio clip of that moment. Raw transcripts are never returned through the feed, the API, or the MCP.

Removal requests are honored.

If an issuer or a speaker asks us to remove a clip or a signal, we remove it from the feed, the API, and the MCP. Reach the team through your account or the request-access flow and we will confirm the removal.