Multi-location service businesses
01
Where it splits
- Same customer, different story on each site — A process that is "roughly the same" is not the same when the clock is ticking. Variation turns into re-work, quality swings, and nobody quite agrees what "done" means.
- Reporting you cannot defend in a Monday meeting — Consolidation in spreadsheets, exports that are already stale, or three definitions of a simple metric. You need numbers that are boring because they are consistent, not because nobody asked hard questions.
- Handovers that do not fit in an email — A site manager leaves, a key person is off, and suddenly three people are on a call to reconstruct what the system should have known. The cost is not just the hour — it is the delay and the risk that follows.
- Another tool one site will not use — Rollouts that work on paper but ignore how each branch really operates. A platform pushed from the top without buy-in, training, and a line on who owns the data, becomes an expensive log-in screen.
02
What we actually do
- Map end-to-end with the ugly bits — Rush jobs, stock fires, who signs what, where HQ helps versus meddles — no diagram that pretends Tuesday does not happen.
- Standard first, stack second — What must match everywhere vs what is allowed to differ — written down so software is not doing office politics by stealth.
- Visible status and owners — One place for job or customer state, clear accountability, audit where someone might actually ask — fewer standing meetings, fewer "I thought they had it."
- Rollout people will use — Pilot on the sceptical site, training in real shift lengths, feedback taken seriously. If the new screen loses to a notebook at 4pm, we fix the flow — not blame the branch.
→
How we work across sites — step by step
One business, not six secret playbooks — standard before stack.
Step 1 Before we start
Book & pick a branch
One site or one job type that shows where HQ and branches diverge — including the exceptions head office pretends are rare.
Step 2 All sites
Map end-to-end with ugly bits
Rush jobs, stock fires, who signs what — no diagram that pretends Tuesday does not happen at every location.
Step 3 Agree the line
Standard first, stack second
What must match everywhere vs what may differ — written down so software is not doing office politics by stealth.
Step 4 People will use it
Pilot, rollout, handover
Start on the sceptical site, training in real shift lengths, feedback taken seriously — fix the flow if the notebook wins at 4pm.
Numbers for Monday that do not need a spreadsheet reconciliation ritual — that is the bar.
04
When it is working
- Sites can describe "how we work" in one sentence — exceptions deliberate, not accidental.
- Numbers for Monday that do not need a spreadsheet reconciliation ritual.
- Handovers that survive holidays and manager churn without a three-hour call.
- Tools managers log into — with maintenance cost you can budget, not discover in year two.
Six branches with six secret playbooks is still one brand to the customer. Book a review — we'll map where work diverges and what must match before anyone sells another rollout.
No single-vendor religion. Core swap, thin custom layer, or integration that kills double entry — whatever matches reality. If the first win is killing a stupid process before more licences, we say that first.