From headcount case to first commit

What are the staffing process steps?

The staffing process steps run in eight stages: build the headcount case, define the role, source candidates, screen them, select through structured interviews, make and close the offer, place the person on the team, and onboard to first productive output. Recruitment is one stage inside that arc rather than a synonym for the whole of it. The two stages that decide whether a hire works sit at the ends, the business case at the front and onboarding at the back, and those are the two most often skipped.

Give every stage an owner and an exit criterion. A stage with no exit criterion is where requisitions quietly stall, because nobody can say whether it is finished.

Two distinctions save a lot of argument later. Recruitment is stages three to six, the market-facing work of finding, assessing and closing people. Selection is stage five on its own, the decision about which candidate clears the bar. Staffing is the system around both, and it starts with a number in a capacity plan and ends with someone producing work.

Why the headcount case comes before the job post

Workforce planning is the step that decides whether the role still exists in three months. A headcount case is a written argument: the work that will not get done without the hire, the delivery or revenue exposed by that gap, the alternatives considered (contract, managed team, automate, do nothing), the fully loaded first year cost, and the date the gap starts to bite. Roles that reach the market without that document are the roles that get frozen in the next budget review, usually after a candidate has already been interviewed twice.

The case is also where the employment model gets decided. Permanent headcount, contract, and augmented delivery capacity are three different processes with three different clocks, and choosing between them after sourcing has started wastes the sourcing.

That number reframes the argument about approvals. A role parked for six weeks in a review queue costs more than the discount anyone negotiates on the recruiting spend.

Workforce planning is not an annual exercise either. Rebuild the picture every quarter against three inputs: the delivery commitments already signed, the attrition you can see coming, and the skills the current team cannot cover without hiring. Roles that fail that test do not need a faster process. They need to be cancelled before anyone spends a screening hour on them.

How long should each of the staffing process steps take?

Benchmark by role level, not against a single company-wide target. Average time to fill for nonexecutive roles sits at 44 days¹, and that average hides a spread of nearly three to one: entry level roles close in about 46 days, manager level roles run 60 to 75 days, nearly 40% of senior positions take more than 90 days, and executive searches average 120 days⁴.

Function matters as much as level. Engineering roles carry the longest median time to hire at 49 days, while administrative and customer service roles close in 33 to 34 days⁴. A single 45 day service level target applied across a whole company is therefore wrong twice over: too slow for the roles that should close inside a month, too aggressive for the ones that never will.

Inside a 44 day nonexecutive fill, the allocation below is a planning target rather than a measured benchmark. It is useful mainly because it forces approval time out into the open:

• Intake and scorecard: 3 days
• Sourcing to first slate: 10 days
• Screening and structured interviews: 12 days
• Debrief and decision: 4 days
• Offer build, approval and acceptance: 7 days
• Buffer for reschedules and second finalists: 8 days

If your real numbers put twenty of those days inside approval steps, a bigger sourcing budget will not fix anything.

Senior roles need the opposite treatment. When close to four in ten of them run past 90 days⁴, the honest plan is a bridge: an interim or contract holder for the work, running in parallel with the search, so that operational pain does not compress the search into a rushed decision at day 70.

Where the staffing process actually breaks

Most staffing failures are not sourcing failures. The candidate was found in week two, the offer then sat in signature loops for three more weeks, and the candidate accepted elsewhere. Sourcing takes the blame because it is the visible stage, and approvals escape scrutiny because no single person owns the queue.

Count the days your process spends waiting for a signature. In most hiring reviews that is the largest single line, and it is the only one you can fix without spending money.

The second break is volume. More applicants did not produce faster hiring, they produced more screening, and screening capacity did not grow to match.

Without a scorecard written before sourcing starts, extra volume is simply extra cost per hire. The scorecard is what makes screening a filter instead of a reading exercise.

The third break is an uncalibrated slate. If the hiring manager and the recruiter have not agreed what a strong candidate looks like on paper, the first slate is a guess, the feedback on it is vague, and the second slate arrives a fortnight later. One calibration call on the first two profiles, held before volume sourcing begins, removes more days than any tooling change.

Interview sprawl is not rigour

Adding rounds is the standard response to a previous bad hire, and it is the wrong one. Structured interviews, where every candidate answers the same questions against the same scorecard and interviewers file scores before the debrief, do more to cut mis-hire risk than a fifth conversation does. Past three or four structured conversations, extra rounds mostly add calendar days and lose the candidates you most want.

Compressing time to fill without dropping the bar

• Pre-approve the requisition and the compensation band before sourcing, not after the finalist appears.
• Book the full interview loop as one calendar block on the day the slate lands.
• Cap the panel at four interviewers and name one decision maker.
• File scores before the debrief so the loudest voice does not set the anchor.
• Keep a second finalist warm until the start date, not until the acceptance.

What the staffing process costs when it drags

Cost per hire is total internal and external recruiting spend in a period divided by the number of hires in that period. It averages $5,475 for nonexecutive roles and $35,879 for executive roles¹. The gap is not seniority alone, it is search time, which is the clearest argument for running executive and volume hiring as separate processes with separate budgets.

The larger cost never appears in the recruiting budget at all. A bad hire costs at least 30% of that employee’s first year expected earnings⁶, and replacing an employee runs between 30% and 400% of salary depending on the role⁷. Set against that range, the marginal spend on a better selection stage is small, and the cost of skipping one is not.

How a staffing partner’s process differs from in-house hiring

A staffing partner runs the same eight stages, but the sourcing stage is largely finished before your requisition exists. That is most of the speed difference. Agencies fill temporary roles in about 6 days, contract roles in about 8 days, and permanent placements in about 32 days⁵, against a nonexecutive market average of 44 days¹.

The trade is control of the channel, not control of the bar. You keep the scorecard, the selection standard and the onboarding; you give up direct ownership of where candidates come from. For engineering roles, where median time to hire is longest⁴, that trade usually pays for surge capacity and rarely pays for the first two or three people on a founding team. Staff augmentation is the version where placed engineers work inside your process and your standards, while hiring dedicated developers is the version where a partner carries a whole delivery unit.

Either way, keep three stages in-house: the headcount case, the scorecard, and the onboarding. Handing those out is how companies end up with fast placements and slow outcomes.

Which metrics prove the staffing process worked?

Time to fill proves the process moved. It does not prove it worked. The metric that does is time to first productive output, the day a new engineer merges a real commit or a staffed consultant closes a first billable week. It is harder to instrument, and it is the only number that connects hiring activity to delivery.

Onboarding belongs inside the staffing process, not after it. The 30, 60 and 90 day reviews are where quality of hire is actually observed: ramp against the scorecard written in stage two, manager confidence, and whether the outcomes named in the headcount case are being delivered. Run those reviews and the loop closes, because a mis-hire becomes a process defect you can trace to a specific stage instead of a story about a candidate.

Track four numbers on every requisition and the bottleneck surfaces within a quarter: days per stage, offer acceptance rate, cost per hire, and 90 day retention against the scorecard. If you want a second read on where your own process is losing weeks, tell us what your last three hires looked like.

Written by the group's editorial team with the practice leads who run these builds. Reviewed before publish. Spotted an error? Tell us and we will fix it.