Time to hire is one number hiding four different problems.
Most TA teams try to fix it by pulling a single lever. More sourcing. Faster scheduling. Better job descriptions. It rarely works. Time to hire isn't one problem. It's four problems stacked on top of each other. Each one needs a different owner, a different target and a different intervention.
Here's a more useful way to look at it.
The four levers
When a hire takes too long, the delay lives in one of four places:
Time to scope. From approved headcount to a recruiter starting work. The JD sign-off. The intake meeting. The comp exception request. The debate about whether this is a senior or a staff role. None of this is sourcing. All of it is lost time before sourcing begins.
Time to source. From first outreach to a qualified candidate in the process. How long does it take to find the right people, get a response and get them through the first screen? This is where recruiter skill and market knowledge actually matter.
Time to interview. From first interview to a hiring decision. Panel availability. Debrief frequency. How quickly the team can align on a yes or a no. This is almost entirely a hiring manager and coordination problem. Recruiters rarely control it, but they can compress it.
Time to close. From verbal offer to signed contract. Counter-offers, slow approvals, candidates in multiple processes. This is where offer strategy and relationship quality determine the outcome.
Most teams collapse all four into a single time-to-fill figure, report it monthly and wonder why nothing changes. The number tells you something is slow. It doesn't tell you where.
Why this distinction matters
Split time to hire into its four components and something useful happens. Each lever becomes measurable and ownable independently.
A team with a 90-day time to fill might have a 5-day time to scope, a 40-day time to source, a 35-day time to interview and a 10-day time to close. Those are four completely different problems.
The scoping is fine. Sourcing is slow, which might mean the role is genuinely niche, the market is thin or the brief is unclear. The interview stage is where the real time is going. Probably because panels are hard to schedule and decisions take two rounds of debrief. Closing is borderline acceptable.
If you try to fix that by pushing the team to source faster, you're optimising the wrong lever.
Breaking a composite metric into its causal inputs is standard practice in most disciplines. In TA, it rarely happens.
Niche roles break the standard model
There's a complication that the time to hire conversation usually ignores. Not all roles respond to the same interventions across each lever.
Volume hiring (early careers, broadly-skilled roles, standardised assessments) can be compressed at every lever simultaneously. Batch intake meetings. Sourcing campaigns running in parallel. Interview days. Cohort offers. The early careers model works because roles are homogeneous. The same signal applies to every candidate. You can run the four levers in parallel because the variables are consistent.
Niche roles are different. A GPU infrastructure engineer, a Principal Distributed Systems Architect, a Founding ML Researcher. These roles have smaller talent pools, more ambiguous briefs and more complex evaluation criteria. The levers can't simply be compressed in parallel. They need to be owned by people with the right knowledge for that specific domain.
Think of it this way. Finding a radio channel on a tuner works best with one person adjusting one dial carefully. Three people adjusting three dials simultaneously adds noise. It doesn't improve signal.
The same principle applies. Putting more sourcers on a specialist role rarely produces better results. It produces more noise in the pipeline and dilutes the signal that tells you which candidates are genuinely strong.
The better approach: put one person with deep domain knowledge on the sourcing lever. Another with process expertise on the interview lever. A senior recruiter who knows the market on the close. Fewer people per lever, higher signal per lever.
What this means for how a TA team should operate
The four-lever model has implications for team structure that most TA leaders haven't fully worked through.
The traditional model is one recruiter, one portfolio of roles, full ownership of all four levers on every req. That model has a ceiling. A recruiter managing 15–20 roles is context-switching across 60–80 active lever problems at once. Some levers get neglected because others are on fire.
A lever-based model distributes the work differently.
Junior recruiters build depth on one or two levers (typically sourcing and interview coordination) across a batch of roles. Senior recruiters operate across all four levers and own the hiring manager relationship. Staff-level specialists own the lever that requires the deepest domain knowledge, particularly for niche roles. Managers set lever targets, distribute work across the team and remove the cross-functional blockers that sit between levers. Slow comp approvals. Unavailable panels. Delayed JD sign-offs.
The distinction between who owns the hiring manager relationship and who executes the lever work matters. In a distributed model, the work can move across the team. The accountability to the business shouldn't.
The practical starting point
Before changing how your team is structured, start with instrumentation.
Pull your last 12 months of closed roles. For each one, identify when the req was approved, when sourcing began, when the first candidate entered the process, when the offer was made and when it was accepted. You now have rough lever timings. They won't be perfect. Most ATS systems aren't set up to capture this cleanly. But they'll be directionally accurate.
Look for where the time is actually going. In most teams, the finding is counterintuitive. Time to scope is longer than anyone thinks. The req approval to intake gap is invisible in the headline metric. Time to interview is longer than it should be. Debrief scheduling gets treated as administrative rather than strategic. Time to source gets the most attention and is often not the primary problem.
Once you know where the days are, you can target the right lever. And once you can target the right lever, you can build a team that owns it properly.
A note on what this framework is and is not
This isn't a new way of thinking about recruitment strategy. It's a decomposition of an existing metric into its component parts. Intake, sourcing, interviewing and offer are not new concepts. The value is in separating them clearly, measuring them independently and designing your team around the fact that they need different skills, different owners and different interventions.
Time to hire will always be a useful headline number. It tells you whether you're competitive. The four levers tell you what to actually do about it.