"Standing Orders Template for Small Ops Teams: The One-Page Order Sheet That Runs the Shift Without You"

#operations #leadership #checklists #devops

Standing Orders Template for Small Ops Teams: The One-Page Order Sheet That Runs the Shift Without You

Companion to the Standing Orders Template page on ops-notes — full sheet, worked example, and related runbooks.

TL;DR

Every small team runs on invisible orders — the things the owner would say if you could ask. Standing orders make them visible: one page that says what "normal" looks like, what you're authorized to do alone, when to wake somebody up, and what never gets touched. If a question comes up twice, it becomes a standing order; if it happens three times, it gets a checklist of its own.

Why Most Small Teams Run Without Them

In the military, standing orders are instructions that remain in force until changed — the sentry doesn't ask what to do if the patrol is late; the answer is already on the card. In small ops teams the equivalent rarely exists. Everything lives in the owner's head: which alerts matter, who can approve a $200 spend, when it's acceptable to reboot the staging box.

The symptoms are easy to recognize:

Affordable decisions wait hours because "I wasn't sure if I was allowed."
The same escalation question gets asked in three chats and answered three ways.
The first incident of the night starts with "I didn't want to wake you, but…"
Everyone's judgment is fine — nobody wrote it down.

The Seven Sections

One page. If it runs past two pages, you're writing a policy manual — cut it back until everything fits on the page someone actually prints.

Watch conditions — expected traffic, batch jobs, and green/yellow/red definitions for the three systems that matter most.
Escalation triggers — exact thresholds, not "use judgment": revenue-affecting outage > 10 min → call; admin-credential security signal → call now; staging down → log it, fix it, mention at handoff.
Delegated authority — spend up to $X per incident and $Y per day; restarts allowed for listed services and forbidden for two named critical ones; template answers allowed, refund promises forbidden.
Comms plan — one-line update every 2 hours in the status channel; who is incident commander when the owner is away; written handoff at a fixed time.
Battle rhythm — shift-start dashboard sweep, mid-shift patch-window check, end-of-shift audit sweep.
Never-do list — raw SQL on prod during an incident; silencing an alert instead of fixing it; credentials over chat.
Day-close checklist — open items logged with owners, handoff written, dashboards green-or-explained, tomorrow's known events listed, sheet date updated.

The Copy-Paste Sheet

STANDING ORDERS — [TEAM/SHIFT]                      Sheet date: ____
Valid until changed. When this sheet and a human disagree, the sheet wins
until the owner says otherwise.

1. WATCH CONDITIONS
   Normal traffic: ____ req/min ± ____   Batch 0230-0310   Dashboards: ____
   GREEN = ____   YELLOW = ____   RED = ____

2. ESCALATION TRIGGERS (call = phone, not chat)
   [ ] Revenue-affecting outage > 10 min .......... CALL OWNER, then IC checklist
   [ ] Admin-credential security signal ........... CALL NOW, page on-call
   [ ] Payment/checkout errors > 5 in 15 min ...... CALL, payment outage playbook
   [ ] Staging down / cosmetic / single worker .... LOG + FIX + note at handoff

3. DELEGATED AUTHORITY (no approval needed)
   Spend: $__ per incident, $__ per day
   Restart allowed: [list]     Restart FORBIDDEN: [two named systems]
   Allowed comms: templates.  Forbidden: refund/credit promises, dates.

4. COMMS PLAN
   Update channel ____ every 2h.  IC when owner away: ____.
   Handoff at ____: written, handoff checklist.

5. BATTLE RHYTHM
   Shift start: dashboard sweep.  Mid-shift: patch-window check.
   Day close: audit sweep + handoff.

6. NEVER-DO LIST
   [ ] Raw SQL on prod during an incident
   [ ] Silence an alert instead of fixing it
   [ ] Credentials over chat

7. DAY-CLOSE
   [ ] Open items logged w/ owners  [ ] Handoff written  [ ] Dashboards green-or-explained
   [ ] Tomorrow's known events listed  [ ] Sheet date updated

Worked Example: 02:14, Checkout Goes Dark

The owner's phone is on do-not-disturb. With standing orders in place, the on-duty tech hits trigger 2 at 02:23 (five payment errors in fifteen minutes) and calls — then opens the payment outage playbook, flips the banner to "payments temporarily unavailable — orders held, not lost," spends $0, promises nothing, and hands off in writing at 06:00. The owner wakes to a resolved incident instead of a missed-call list. Without standing orders, the same night starts at 09:30 with a support inbox on fire.

Keeping the Sheet Alive

Write v1 in 30 minutes — done beats perfect.
Rule of two: every question asked twice in a week becomes a line on the sheet.
Review at handoff, weekly — crossed-out triggers and new dollar limits keep it honest.
Version it — sheet date at the top, one page, printed copy next to the runbook binder.
Test it — one tabletop drill per quarter: "owner unreachable, checkout down, go."

Daily Ops Audit Checklist — the day-close sweep as a daily habit
Incident Commander Checklist — who runs the incident when the sheet says call
Escalation Policy Template — the full ladder behind trigger section 2

Field-Tested Ops Kits

This template is part of the Hive80 Lab ops kit line — instantly downloadable: The First 30 Minutes (free), Ops Field Cards ($4), Ops Starter Kit ($14), Ops Mega Bundle ($29). Launch pricing is live at the store — check the banner for today's code.

DEV Community