# Time to slate: the only hiring metric that moves.

_Why screen-ready candidates within hours of a role going live beats every vanity metric in your ATS._

Published 2026-11-10 · Updated 2026-11-13

Most talent teams report time to hire. Boards ask for it. ATS dashboards headline it. It measures the wrong thing.

Time to hire starts when someone is already in process. It ignores the slow weeks before a hiring manager sees a name they would happily interview. It ignores the false starts: searches that never become slates, slates that never survive first pushback from a leader. By the time the clock starts, you have often already lost the candidates who would have moved.

Time to slate is different. It is the interval from a role going live to a shortlist your recruiters can defend: names, reasons, evidence, timing. Not a CSV export. Not thirty profiles in a shared search. A slate a hiring manager can act on this week.

When slates arrive in minutes rather than weeks, downstream metrics move without magic. Time to first interview drops because nobody is waiting for a manual search. Offer acceptance improves because you contacted people while they were still open. Hiring manager confidence rises because every name comes with a sentence they can repeat in a staff meeting.

Most teams underestimate how much delay is search latency, not process. The briefing call is on Tuesday. Legal approves the level on Thursday. The role posts Friday. Then recruiting spends eleven days rebuilding a search everyone has run before, pasting Boolean strings, chasing the same brands. The manager pings on Wednesday asking if AI can speed it up. You send ten more lookalikes from the same pool.

Agent mode changes the starting point. The moment a role is assigned in your ATS, a people world model can read the team, draft the brief against your culture profile, scan the market for undervalued fits, and return a screen-ready slate while the requisition is still new. The recruiter edits, challenges, adds humanity. They do not restart from zero.

That only works if the model knows your company. Shared AI sourcing cannot compress time to slate in a durable way. It returns the same ranked list as last month with fresher adjectives. A private model, trained on your outcomes, gets sharper each quarter. First weeks it learns teams and strategy. First quarter it learns who succeeds here. Every hire and every rejection teaches it, and that teaching stays yours.

Measure what you can defend. Track time to slate alongside time to first interview. When slates slip, Orbit flags it before the manager escalates. When a finalist stalls, you see whether the bottleneck is scheduling, evidence, or a weak reason. Compare finalists signal by signal instead of debating a single score nobody trusts.

Vanity metrics comfort dashboards. Time to slate changes whether your leaders believe recruiting is a constraint or a partner. If the only number you report is time to hire, you are optimising paperwork. If you report time to slate, you are optimising judgement at the moment it matters.

If you are rebuilding hiring operations this year, put time to slate on the exec deck. Define it the same way everywhere: live role to defendable shortlist. Then ask whether your stack can produce that slate without re-running the same market search your competitors run.

Related reading: why language models alone cannot deliver that slate, how Orbit defines signals instead of scores, and our security approach to keeping interview data private per customer.
