Workforce Architecture

Composite example · Architecture → Intervention → Result

“We cannot find people” is almost never the problem

A precision manufacturer under institutional ownership needed to double direct-labor capacity. Every conversation began with the wage. The read found the constraint somewhere else entirely — and the actions that resolved it cost almost nothing.

This page answers one question: When an enterprise says it cannot hire, where is the economic truth — and what does the operating system do with it?

Step 01: The problem as stated

What everyone agreed on before the read.

The stated problem

“We cannot find machine operators.” Every conversation started and ended there.

What the numbers said

The requisition count was not the binding constraint. Time-to-productive was. Hires arrived and then waited on setup availability, so each new operator consumed senior-operator hours before returning any.

Where the loss was booked

The loss did not appear in a recruiting line. It appeared as overtime, expedite freight, and jobs held for the three people qualified to set up the constrained cells.

The single number that reframed it

Cost per productive operator — fully loaded through the first ninety days — was several multiples of the posted wage gap the team was arguing about.

Step 02: The labor geography

Before any process work: what the region can actually supply, and at what cost.

Three commute rings, each with a different supply, a different cost to reach and a different retention behavior. Select a ring or an institution to open its record.

Pipeline institutions

0–25 minutes

The dependable core

The same three rings, on the ground

The drawn figure holds the logic; the map holds the geography. Both read from one record.

0–25 minutes · within about 13 miles

The dependable core

Every conversation began with the wage, so the wage is worth reading properly. Move the selector and watch the only two figures that decide capacity move with it.

Wage position selector

Where the enterprise chooses to sit in the local market

50 th percentile

Hourly rate

$19.05 per hour

The reference position — the market midpoint.

Days to first start

47 elapsed days

Offer acceptance 58% at this position.

12-month turnover risk

34.0% separations

The reference separation rate.

The market midpoint. This is the reference position every other scenario is measured against.

Wage anchors: BLS May 2023 Dayton–Kettering, OH MSA (19430), operator family, uprated 11% cumulatively to a 2026 planning basis on an ECI trajectory. Acceptance, elapsed days and the 12-month separation rate are modeled planning elasticities, not measured facts — one completed hiring cycle replaces them with actuals.

The wage is only one line of the bill. Read the premium, the open seats and the leavers in one total, and the cheaper path is rarely the lower rate.

Sensitivity — total annual sourcing cost

What a hiring year costs when wage premium and time to fill are read together

Each cell is one year of total sourcing cost for 100 hires: the wage premium above the market midpoint, the contribution lost while seats stay open, and the replacement of the cohort that separates inside twelve months. Rows are the wage position; columns are the process running faster or slower than the modeled days.

Lighter = lower total Darker = higher total

The reading

Across the modeled range, time to fill moves total annual sourcing cost by $882K and wage position by $550K. The cheapest cell on the grid is the 50th percentile at −20% days ($2.07M); the most expensive is the 90th at +25% days ($3.55M). Paying a premium is only the cheaper path when the days and the leavers it removes are counted in the same total.

Cell detail

Select any cell to open its three cost lines and its difference against the midpoint at the modeled days.

Basis: BLS May 2023 Dayton–Kettering, OH MSA (19430) operator wage percentiles, uprated 11% cumulatively to a 2026 planning basis; 2,080 straight-time hours. Acceptance, days to fill and 12-month separation are modeled planning elasticities, replaced by the Post-Hire Worksheet actuals once one hiring cycle is measured.

Step 03: The brown paper

The current-state process on the wall, drawn by the people who run it.

• 01 Requisition to offer ran through nine hand-offs across four functions, and no one owned the whole line.
• 02 Two approval steps existed only to confirm an approval already given upstream.
• 03 The written job requirements had not been rewritten in years; they screened out the exact population the region actually produces.
• 04 Background and drug screening ran sequentially rather than in parallel, adding days for no added assurance.
• 05 Onboarding began on a fixed calendar date, so a candidate who accepted on day two waited up to three weeks to start.
• 06 No step in the process created a trained setter — the constraint that everything else queued behind.

Step 04: SIPOC of the hiring line

The process boundary agreed before anything is improved — supplier to customer, with the measure each step is held to.

The brown paper shows what happens. The SIPOC states what the line is: where it starts, where it ends, who supplies it, who receives it, and one owner accountable across all of it. Without this boundary, every improvement argument becomes a departmental one.

Boundary — starts

A production plan change creates a validated direct-labor requirement.

Boundary — ends

The hire is productive at standard on a qualified machine and retained at day 90.

Owner

One value stream owner for the hiring line (not recruiting, not the plant alone).

• Production planning
• Value stream owner
• Finance
• Demand plan and takt
• Current setter and operator capacity
• Approved wage band

Define the requirement

• Written requirement with skill floor and start date
• Talent acquisition
• Hiring manager
• Technical colleges and workforce centers
• Referring employees
• Job boards
• Role posting written to the skill floor
• Wage position vs market percentile

Attract and source

• Qualified applicant flow
• Screening
• Assessment instruments
• Written job requirements
• Structured interview guide and scorecard

Screen and assess

• Ranked, evidenced shortlist
• Interview panel
• Compensation
• Scorecard evidence
• Wage band and premium rules

Decide and offer

• Signed offer with a firm start date
• Candidate
• Plant scheduler
• HR onboarding
• Safety
• Cell lead
• I-9 / W-4 and background clearance
• Machine and safety curriculum

Onboard to standard

• Cleared, badged, safety-qualified new hire
• Setter
• Setters
• Training pathway
• Setup capacity
• Sign-off standard per machine
• Career ladder

Qualify and retain

• Operator productive at standard, retained at day 90
• Production
• The customer of the product

Process 01

Define the requirement

Process 02

Attract and source

Process 03

Screen and assess

Process 04

Decide and offer

Process 05

Onboard to standard

Process 06

Qualify and retain

Measure pair for the whole line: Lead time (requirement to productive) paired with quality (day-90 retention at standard).

Step 05: Non-value-add and bottlenecks

What the process consumed without adding value, and what everything queued behind.

Non-value-add

Duplicate approvals and re-keyed candidate data

Days of elapsed time and two systems that disagreed about the same candidate.

Sequential screening and paperwork collection

Offer-to-start stretched past the window in which candidates stay available.

Requirements that over-specified experience

The largest local supply — trained entrants — was filtered out before a human read the application.

Bottleneck

Setup-capable operators

The true constraint. Three people gated schedule, training and quality response for the whole plant.

Fixed-date onboarding

A queue of accepted candidates waiting, with the highest drop-out rate in the process.

One person owning both sourcing and interviewing

Throughput capped by a single calendar regardless of demand.

Step 06: SWOT of the process

Not a SWOT of the company — a SWOT of the hiring line itself.

Strengths

• Real technical work that skilled people want to do.
• A dependable workforce core living inside twenty-five minutes.
• Leaders who knew the shop cold and answered questions with facts.

Weaknesses

• No owner of the hiring line end to end.
• No internal progression path, so the only way up was out.
• Requirements written for the market of a decade ago.

Opportunities

• Two nearby institutions able to supply a scheduled intake rather than a scramble.
• A registered apprenticeship structure already available to carry the internal ladder.
• Parallelizing screening removes days without spending a dollar.

Threats

• Larger employers on the same corridor setting the wage.
• Knowledge concentrated in three retirement-eligible operators.
• Wage inflation absorbing the growth thesis if the answer stays “pay more, farther out.”

Step 07: Funnel and conversion arithmetic

Edit the assumptions. Offers, starts and 90-day retained hires are computed per channel against the target.

A hiring plan is a conversion claim. Enter the volume each channel can actually put into the line and the rate at which each stage passes candidates through; the model states what that produces in retained hires and where the mix falls short. When the number is short, the answer is usually the weakest conversion, not more volume.

The number the plan has to deliver in this period.

Top of funnel required

1,144

Sourced candidates needed at the current blended conversion of 4.4%.

Plan against target

− 16

The channel mix is short of the per-channel targets. Fix the weakest conversion before adding volume.

Assumptions only — nothing is saved. Replace them with your own stage history before the plan is committed.

Step 08: Prioritized actions

Ranked by what has to be true first. Each one carries an owner, a horizon and a measure.

Step 09: Architecture → Intervention → Result

The record of what changed, stated without claims we cannot evidence.

Architecture

The hiring line was treated as a value stream with one owner, a measured lead time and a named constraint — not as a recruiting activity.

Intervention

Economic truth pass, brown paper of the current state, NVA and bottleneck identification, SWOT of the process, and a ranked action set with owners and measures.

Result

The argument moved off wage and onto lead time and setter capacity. The measures the enterprise now reviews weekly — time-to-productive, offer-to-start, setup-qualified count — did not exist before the read.

Evidence tier: composite example. This record is assembled from the pattern of the work and is de-identified — no client, city or institution is named. Verified client case studies and anonymous client stories are labelled separately.