USA
AU
UK
ES
FR
BR
TH
INHow to Choose a Shift Pattern: A Decision Framework
There are dozens of named shift patterns and they all come with enthusiastic advocates. The good news: you don't need to evaluate dozens. Five questions, answered in order, narrow the field to two or three candidates — and the final pick belongs to your team, not to a spreadsheet.
1. What hours must you actually cover?
Not 'we're basically 24/7' — chart it. Demand by hour, by weekday, for a typical month. Operations discover they're really 6am-10pm businesses, or that Sunday nights run at 20% of Tuesday's volume. The answer sorts you into one of three families: true 24/7 (the 24/7 models), extended/7-day (two shifts, staggered or weekend designs), or weekday/compressed (4/10, 9/80).
2. What shift length suits the work?
This is the 8 vs 10 vs 12 decision. The short version: heavy or high-vigilance work favours 8s; days-off-hungry teams in lighter work favour 12s; peaked demand favours 10s with overlap. Length picks your candidate patterns' division.
3. Fixed crews or rotation?
If genuine night-shift volunteers exist in your labour pool, fixed designs (Pitman) harvest that preference. If nights would be conscription, rotate and share them (Panama, DuPont). The full argument is in fixed vs rotating shifts — and if you rotate, direction and speed are non-negotiable hygiene (forward, fast or properly slow).
4. What break shape does your team value?
Patterns with identical weekly hours distribute them in wildly different shapes. The same 42-hour average buys: every other weekend off (2-2-3 family) · a 7-day break every month (DuPont) · 4-day breaks every 8 days (4 on 4 off) · whole alternating weeks (7 on 7 off). None is objectively better — ask which one your actual team would defend.
5. Can you staff it?
Every 24/7 pattern needs roughly four crews plus 10-15% relief (the coverage math). If the headcount isn't there, don't run a four-crew pattern on three crews — shrink the covered hours or fund the hiring first. Chronic structural overtime corrodes every pattern equally.
Then: shortlist, pilot, vote
Take the two or three survivors to the team with honest one-page summaries (the pattern library pages are built for exactly this). Pilot the favourite for one full cycle — a pattern is only knowable lived — then vote. A team that chose its roster will make almost any reasonable pattern work; a perfect pattern imposed will fail on contact with the first bad month.
Build this schedule in Tommy
Set the rotation once and Tommy fills the weeks ahead — shift swaps, leave and coverage gaps handled in one place, with your team always seeing the latest version.
