If you've watched how the more heavy-handed outreach tools operate, you've probably seen the failure mode: someone applying for a mid-level individual-contributor role gets a VP or a founder cold-messaged on their behalf. It reads as either a mistake or an imposition, and either way, it doesn't help.
Good targeting isn't about reaching the most senior person you can find. It's about reaching the person actually positioned to act.
The order, and why it's in that order
Day 1 — HR or the recruiter. Whoever is named on the job post, or whoever at the company has a bio that says they're hiring for that function, gets contacted first. This is the least presumptuous move available: recruiters expect and want this kind of outreach, and they're the person with an actual process for moving a referral forward.
Day 2 — A peer, always. Regardless of what happened on Day 1, a peer gets contacted too — someone doing a similar job, at a similar level, ideally in the same location. Not a fresher, not someone three levels up. A peer can speak to what the role is actually like day to day, which carries a kind of credibility a recruiter conversation doesn't.
Day 3 — A follow-up, if neither has responded. This is where most real replies land — a lot of people mean to respond to the first message and just don't get to it.
Day 4 — Someone more senior, only if nothing has worked by this point. This is the fallback, not the opener.
Why the order matters, not just the people
Sending all three at once looks efficient but defeats the purpose. A recruiter who sees you've also messaged three of their colleagues in the same hour reads that as automation, not intent. Staging the outreach — and stopping the instant anyone replies — keeps each message looking like what it actually is: a specific, considered attempt to reach the right person, not a blast.
What happens when the waterfall comes up empty
Sometimes a company has no visible HR presence and no obvious peer to reach — small teams, sparse LinkedIn footprints, unusual org structures. When that happens, the honest answer is to say so clearly rather than force a match: we surface that this company doesn't have a resolvable contact yet, rather than sending a message to whoever happened to be closest to the criteria. A best-effort message to the wrong person isn't better than no message. It's often worse, because it's the version of outreach that actually does read as random.
The point of doing it this way
None of this is about being clever with sequencing for its own sake. It's about the fact that a referral request is, underneath the automation, still a favor you're asking a stranger to extend. The staging exists to make sure that favor is being asked of someone who's actually in a position — and reasonably likely to be willing — to grant it.