# First 90 Days — Onboarding Cycle Guide

> Folder: `first-90-days` · Category: scenario kits · 1 new hire + manager/HR pair · 30 min per checkpoint · offline + online · print props · scenario kit
> License: CC BY 4.0

New hires decide whether they made the right choice four times: on day one, and then at thirty, sixty and ninety days. Most companies only hold the first meeting and hope through the rest. This kit runs the whole cycle — a day-one welcome sheet that fits on one page, three checkpoint conversations with written questions, a confidence pulse the new hire fills in anonymously, and a risk watch-out that names the quiet exits before they happen. For HR teams, hiring managers and any small company that cannot afford to lose a hire in month four.

## Objectives

By the end of the ninety days, the pair (new hire + manager) will have:

1. A shared written picture of success: what the first week, the first month and day ninety each look like — signed by both, not assumed.
2. Three checkpoint conversations that happened on schedule, each ending with two things kept and one thing changed — recorded where the next checkpoint can read it.
3. A confidence trend from the pulse surveys (week 1 / 30 / 60 / 90) that tells you who is quietly disengaging while their output still looks fine.

## When to use this kit

- You hire one person at a time and "onboarding" currently means a laptop, a wiki link and good intentions.
- A new hire left at month three and everyone was surprised — the signals were there, nobody had a place to write them down.
- You are an HR team standardizing onboarding across managers whose styles range from daily coffee to "ping me if lost".

## What this is not

- Not a compliance checklist. Contracts, payroll and legal paperwork belong to your HR system; this kit runs the human cycle around them.
- Not a performance management tool. The checkpoints are two-way — the new hire grades the onboarding too — and none of the pulse answers feed a review.
- Not a retention insurance policy. A bad role fit surfaces by day 30; the kit's job is to surface it, not to hide it with perks.

## Setup

**Space:** the checkpoints are sit-downs (or video calls) of 30 minutes; day one is a desk-side conversation of 20.

**Materials per new hire:** one day-one welcome sheet (printed or sent before start), one expectations card signed by both, the pulse survey link or printed slips for weeks 1/30/60/90.

**Materials for the manager:** the checkpoint question cards — you ask from the cards, not from memory; the cycle review sheet for your own notes between checkpoints.

**Grouping:** one pair per cycle; HR runs the same deck across all pairs and reads the aggregate pulse at day 90.

**Roles:** the manager owns the schedule — a checkpoint that slips a week is a checkpoint that says what the company really prioritizes. The new hire owns the honesty; the anonymity of the pulse is what protects it.

## How to run it

| Time | Phase | What you do | The new hire does |
|---|---|---|---|
| Day 1 | Landing | Walk the welcome sheet: week one, first month, day ninety — in writing, both sign the expectations card. Answer the question they are too polite to ask: "how do I know I'm doing well?" | Fills the week-1 pulse anonymously; writes one thing they were afraid to ask. |
| Day 30 | First real read | Run the checkpoint questions from the cards. Two kept, one changed — written down. Compare the pulse trend against your own read. | Rates the onboarding as honestly as the job; names the one thing still unclear. |
| Day 60 | Course correction | Same loop, deeper: what's the hardest part now? Check the changed item from day 30 — kept or drifted? | Sees their day-30 answers acted on or hears why not — either is fine, silence is not. |
| Day 90 | The account | Close the cycle: read all three sheets together. Keep/stop/start on the role itself. Decide and write: what does the next quarter's goal look like? | Gives the honest review of onboarding — the one input future hires will benefit from. |

## The card deck

Use the faces from `first-90-days-prop-cards.pdf`: a master card (the cycle in one view), the day-1 landing card, the day-30 and day-60 checkpoint cards, the day-90 account card, and a watch-out card (the five quiet signals of a month-four exit) — plus branded backs and blanks for your own role-specific questions.

## Failure modes

1. **The checkpoint that slips.** Week one becomes "we'll do it after this sprint" and the cycle dies politely. Rescue: all three checkpoints go on the calendar on day one, or they do not exist.
2. **Manager monologue.** The checkpoint becomes a status report; the new hire performs confidence. Rescue: the cards front-load the questions the new hire asks the manager — half the page is theirs.
3. **The pulse nobody reads.** Surveys collected, trends ignored — worse than no survey, because it teaches that honesty changes nothing. Rescue: the day-90 sheet has a line: "what the aggregate pulse told us, and what we changed."
4. **Optimism bias at day 30.** Output looks fine, the person is already gone in their head. Rescue: the watch-out card is read aloud at every checkpoint — quietness, cancelled 1:1s, no questions, vague answers about friends at work.

## Variations

- **Remote-first:** the pulse is a form (the folder's FormLM block), the day-1 sheet is a shared doc; the cards still print — a mailed welcome envelope with the deck lands better than a PDF.
- **Small team, no HR:** the manager runs the cycle solo; the cycle review sheet is the handoff document when you finally hire that first HR person.
- **Interns and seasonal:** compress to day 1 / day 15 / day 45; the questions survive the change of scale.
- **The other direction:** run the day-90 review questions on the manager too — onboarding is being graded whether you ask or not.

## Facilitation notes

- Say the retention truth out loud on day one: "if this is wrong for you, telling us early is a favor, not a failure." It is on the card; most managers never say it.
- The signed expectations card is the artifact that prevents the month-four surprise. Keep it in the same folder as the pulse slips.
- A checkpoint that surfaces a bad fit is a success, not a failure — the ninety-day cycle exists to find out, not to perform success.
- Read the aggregate before you read the individual: five people drifting at day 60 is an onboarding bug, not five weak hires.

## Put it on the wall

Print `first-90-days-activity-poster.pdf` (A3): the cycle, the four checkpoints, the quiet signals — for the HR wall, and photograph it into the onboarding doc.

## Props in this folder

- `first-90-days-prop-cards.pdf` — poker-spec print sheet (63.5 × 88.9 mm, bleed and cut guides): master, day-1 landing, day-30, day-60, day-90, watch-out, branded backs and blanks
- `first-90-days-run-sheet.pdf` — the print edition: this guide + the cycle review, bound as one A4 document with bookmarks (`-Letter.pdf` is US Letter)
- `first-90-days-activity-poster.pdf` — the whole cycle on one A3 wall sheet
- `first-90-days-guide.md` · `-debrief.md` · `-start-here.md` — the cycle guide, the review sheet, the one-page start here
- `first-90-days-license.md` — what you may do
- `SKILL.md` · `first-90-days.zip` — the machine-readable manifest, and the whole folder in one download

## Agent version

Every file in this folder has a stable public URL — an AI agent can fetch the card deck, the review sheet and this guide with three curl commands (see `SKILL.md`), or build the pulse survey from the week-1/30/60/90 items. The paper version needs two calendars that agree. Keep the pulse anonymous; a named confidence score is a performance review wearing a mask.