PCS Veteran Onboarding: 30/60/90-Day Artifact Checkpoints, Not an 82% Claim

This guide gives a 30/60/90-day onboarding plan for veteran hires after a PCS built on dated checkpoints and presentable artifacts.

It also separates the uncorroborated 82% retention claim from results measured in your own matched cohort.

How the PCS onboarding clock really runs

A PCS start does not create one onboarding calendar; it creates three desynchronised clocks. The employer clock starts on the first payroll day, the household-goods clock runs on the servicing transportation office's own shipment window, and the credential clock runs on the state licensing board's published processing page. Each clock must be verified on its source page, not from the hire's memory or a recruiter's summary, because a PCS move makes all three easy to misremember or compress into one "start date."

Trainual's 30-60-90 onboarding template names a failure mode this desync amplifies: a plan built for a single first-quarter arc, not for a hire whose first weeks are consumed by outside clocks. To check this against your own PCS starts, review the Trainual template before day 1 and identify which entries assume shipment, licensing, and payroll are already settled. In a PCS start, the day-one schedule is often consumed by shipment tracking, licensing paperwork, and payroll setup before role work begins, so any week-one gap in the template arrives even sooner; verify on the Trainual source page which phase actually applies to your hire.

The desync rule is this: a PCS start is desynchronised when the three clocks do not share one verified start date, and the onboarding plan must be written to the latest of the three verified starts, not to the employer's payroll date. The payroll date can open the employer clock, but it cannot certify that household goods or credentials have started.

Practically, that means three separate checks before any checkpoint is treated as a normal deliverable review: confirm the first payroll day from the employer's payroll source, confirm the shipment window from the servicing transportation office, and confirm processing status from the state licensing board's published page. If one clock is still pending, the checkpoint should be used to verify that clock's status rather than to judge the hire's role output.

This is why a PCS hire's first 90 days need dated checkpoints with finished artifacts: the dates keep the three clocks visible, while the artifacts show whether the hire has moved past clock-driven setup into actual role work. The checkpoint is not a support calendar; it is the point at which the hire demonstrates that the slowest clock no longer controls the week.

What six onboarding sources actually prove

Across the six reviewed onboarding sources, the convergence is procedural rather than statistical: each treats the first quarter as a sequence of phased milestones, and the shared claim is that milestones make the first quarter legible. The unsupported leap is that any one milestone design has been shown to produce a specific retention rate; none of the reviewed pieces attaches a retention percentage to that structure.

Keka Academy, published 14 July and last modified 8 October, states that a 30-60-90 plan divides the first three months into focused milestones. That is the verifiable part of the piece. Any further claim that structured onboarding produces higher engagement, productivity, or retention should be checked directly against the Keka page before quoting it: as reviewed here, the source attaches no rate, cohort, or measurement to its framework.

Formester's guide supplies the only hard interval in the reviewed set: checkpoints of 60–90 minutes each, longer than the weekly 1:1, at which the hire presents what they have done and learned and the manager assesses progress against the milestones.

The other reviewed sources describe templates, checklists, problem lists, or AI-assisted variants of the same framework; they do not add a second hard checkpoint interval or a retention rate to the record.

That leaves a clear boundary between what is sourced and what is not. Keka supports the claim that structured onboarding is associated with better engagement, productivity, and retention in general terms; Formester supports a specific checkpoint length and a presentation-and-assessment format. Neither supplies a retention rate, a cohort definition, or evidence that a particular checkpoint design causes retention.

For a PCS program, the transferable element is therefore the checkpoint mechanics, not a headline percentage: hold dated checkpoints long enough for a real presentation, require the hire to show finished work, and measure success against your own matched cohort rather than a borrowed figure. That keeps the program's evidence tied to its own population and prevents a qualitative source from being quoted as a retention guarantee.

Three plan designs compared: pick one

Choose among the three designs by asking what can be shown at the checkpoint, not by how tidy the calendar looks. Swiftools' test is the tiebreaker: each milestone must be a deliverable, not a feeling. A date on a calendar is not a deliverable, and a completed arrival checklist is not a 90-day plan.

Option A is a day-one orientation checklist. It covers paperwork and building access, then stops. Trainual's description of the schedule that fills day one and runs out by the end of week one fits this design exactly. Because nothing is produced after that first week, there is nothing to score at day 30.

Option B is a date-only 30/60/90. The calendar exists, but nothing is produced. It fails Swiftools' test because no one can show the checkpoint anything on the day; the milestone is a date rather than a deliverable, so the plan cannot be assessed.

Option C is a deliverable-based 30/60/90. It holds three checkpoints of 60-90 minutes, and at each one the hire presents what they have done and learned while the manager assesses against the milestones. The checkpoint is not complete unless a finished, reviewable artifact is on the table.

The practical difference is scoring. A checkpoint with an artifact can be marked complete, incomplete, or needs revision; a checkpoint with only a date can only be marked attended.

That is why a date-only plan survives a review but fails at the checkpoint itself.

Pick Option C. Keep Option A's paperwork and access work as the first-week input, and reject Option B's calendar-only structure. Report program success only as your own matched cohort, not as an unverified headline percentage.

Costs: manager minutes and the wrong-hire bill

The manager-time cost of one PCS hire's first 90 days is 570–660 minutes, or 9.5–11.0 hours, when the plan uses three dated checkpoints plus a weekly 1:1. Hold the checkpoint length at the sourced 60–90 minutes; do not compress it to fit a support calendar. The 30-minute 1:1 length is a stated planning input here, not a sourced standard, so replace it with your team's actual 1:1 length before using the total.

The PCS deliverable card

The only new artifact in this section is the copy-usable PCS 30/60/90 Deliverable Card, with a 0/1/2 scoring column. Set the start date on the first payroll day, then fix checkpoint dates at start plus 30, 60, and 90 calendar days before day 1. Each checkpoint runs 60–90 minutes (Formester), while the weekly 1:1 stays at your standard length. Fixing both the dates and the minutes in advance means a shipment delay cannot move them.

Winner for this example: the artifact-gated 30/60/90 design. Break-even trigger: the hire must present a finished artifact at Day 30, Day 60, and Day 90; the first checkpoint without one is where this example stops earning success credit. The per-hire credit is only an input to a later matched-cohort comparison, and this example does not convert credits into a retention percentage.

Decision rules for the first 90 days

Run the first 90 days as three fixed checkpoints—day 30, day 60, and day 90—each held for the sourced 60–90 minutes, with the hire presenting a finished artifact at every one. If a milestone cannot be shown as an artifact the hire presents, it is not a milestone under the Swiftools test; rewrite it before day 1, not on the checkpoint day. A checkpoint is not a support meeting: the date stays fixed even when the work behind it is incomplete.

If the household-goods shipment or state credential is unresolved at day 30, keep the checkpoint date and downgrade the day-30 deliverable to access-plus-artifact. "Access-plus-artifact" means the hire shows what access, systems, or permissions are already in place plus one small finished item that does not depend on the unresolved item; the checkpoint itself is never canceled or postponed.

At each checkpoint, the hire presents first and the manager assesses against the milestone second. If the artifact is missing, record the miss as a dated finding in that same 60–90 minute session and assign the rewrite deadline before the next checkpoint; do not convert the session into coaching or move it to another date.

Credit veteran status only under a matched-cohort condition: if a PCS hire's 12-month retention beats a matched cohort on role, site, prior experience, and start window, report the program as successful for that cohort. If it does not beat that cohort, report the result as unproven and do not attribute the outcome to veteran status.

Apply three final checks before reporting results: the checkpoint happened on its date, the hire presented an artifact, and the comparison cohort was matched on role, site, prior experience, and start window. If any check fails, the program has not earned the credit, even if the hire is retained; do not substitute an uncorroborated headline retention figure for the matched-cohort comparison.

Frequently Asked Questions

What starts the employer onboarding clock after a PCS?

The employer clock starts on the first payroll day.

Whose shipment window governs the household-goods clock?

The household-goods clock runs on the servicing transportation office's own shipment window.

Where is the credential clock tracked?

The credential clock runs on the state licensing board's published processing page.

How should each of the three clocks be verified?

Each clock must be verified on its source page, not from the hire's memory or a recruiter's summary.

What does the hire present instead of following a support calendar?

The hire presents a finished artifact at each checkpoint rather than following a support calendar.

How is onboarding success reported in this model?

Success is reported only as your own matched-cohort result.

Quick answers

Also worth reading: What a connected veteran workforce strategy looks like in practice: What a connected veteran workforce · Veteran job retention data: 2026 SkillBridge 82% vs 69% registered apprenticeship: Veteran job retention data: 2026