Connect, match, credit.
Sygnal reads the tools you already run. There is no tag to add to your site, no change to how your team works, and nothing is written back to your systems.
Connect your stack
Authorise each tool once. Your CRM is first, because it carries the identity everything else resolves against. We read on a schedule and nothing is written back.
The first sync backfills four hundred days and resumes from where it stopped if it fails. A connection that has been failing shows on the dashboard, because stale attribution presented as current is worse than none.
Match touches to accounts
Every touch is resolved to a person and a company, so a LinkedIn view and a webinar registration land on the same account timeline.
Matching is deterministic, on email domain, with public domains excluded. The ones that do not match are not dropped, and not guessed at either. They go to a queue for you to resolve, which matters most on event data, where personal addresses are normal.
Read credited pipeline
Pipeline and revenue credited across the whole journey, under whichever of the four models you want to look through, down to the named companies in each number.
No single model is the answer, and Sygnal does not pretend to pick one. The spread between them is the finding.
What Sygnal does not do
It does not prove causation
Incrementality here is an observational comparison against accounts never exposed to an event. A randomised holdout would be the stronger answer and it is a much bigger piece of work.
It only covers events you host
Your own events, webinars and dinners, where you have the guest list. Sponsored booths and conferences you merely attend are not measurable this way, and nothing in Sygnal implies otherwise.
Podcast attribution is company level
Listener data does not resolve to a person the way a webinar registration does. It is directional and it needs a subscription on your side.
Bring your channel mix.
Twenty minutes, screen shared, and you can pick the channel you are least sure about.