What it is
Signal corroboration is the requirement that a single buying signal be confirmed by at least one independent source before it is allowed to trigger outreach. A funding-round mention on its own is a claim. A funding-round mention that also shows up as a headcount spike, a new leadership hire or a tech-stack change on the same account is corroborated: two unrelated pieces of evidence now agree, which makes the underlying story far harder to be wrong about.
The clearest case is the job posting. A live vacancy on a careers page looks like unambiguous intent, but on its own it is unverified: it might be stale, speculative, or a ghost job with no real budget behind it. Corroboration means checking that posting against something independent before treating it as real, a funding event, a headcount trend, or the account's own track record of postings that actually converted, rather than reaching out the moment the listing appears.
A signal that cannot be corroborated does not get to spend a consultant's touch.
Why it matters
A consultant has a limited number of credible touches with any given prospect, and a badly timed one costs more than the wasted minute. Reach out on a signal that turns out to be false and you burn a touch, look uninformed to the prospect, and make the next message from your agency harder to land. Multiply that across a desk chasing every raw job posting or headline that comes through, and a meaningful share of postings carry no real hiring intent at all, so treating a single posting as sufficient reason to act guarantees some of those touches are wasted before the conversation even starts.
The compounding cost is trust in the feed itself. Once a consultant has been burned by two or three signals that led nowhere, they start discounting every alert that follows, including the genuine ones, which is exactly how a useful signal ends up ignored. Corroboration is what keeps a signal-led desk from sliding into that state: it is not extra caution for its own sake, it is what keeps the feed worth trusting.
How boilr handles it
boilr never lets a single detection stand alone as proof of intent. Before any signal becomes a task, your AI sales employee checks it against what else it knows about the account: other signal types firing in the same window, how fresh and specific the evidence is, and the account's own history in the Company Brain, including whether its past postings or announcements actually turned into real mandates. A job posting is the clearest example, boilr corroborates every posting against other signals before it triggers outreach, but the same gate applies across all twelve signal types it watches.
Corroboration is a pass or fail check, not a ranking. A signal either clears the bar or it does not, and one that does not clear it is held back rather than discarded, ready to be reconsidered the moment a second source appears. That is distinct from signal stacking, which weighs how many corroborated signals an account has already accumulated to decide how urgently to act. Corroboration decides whether a signal earns a task at all; stacking decides how loudly it should shout once it has.
