Dispatch → QuickBooks sync
HVAC, 18 trucks. ServiceTitan dispatch to QuickBooks Online. Manual re-entry at 25 hours/week. APIs stable.
- ROI · 12:1 inside year one
- Complexity · 2 systems, REST
- Stability · both vendors mature
Stop automating the wrong things
The Automation Decision Framework.
Three axes, one matrix. We score every Quick Win on ROI, complexity, and technical stability before we touch a keyboard. The math is on the page in writing. If your project isn't worth automating, we say so. This is the same framework we put through every audit, applied to your operation.
The three axes
Three numbers we score.
Every Quick Win on the audit gets scored on three axes. Anything that lands in the green column on all three is a Quick Win. Anything red on any axis is a no-build.
Axis 01 · Return
ROI
Hours saved times frequency, valued at a defensible hourly rate, weighed against the build and maintenance burden. We require a 3:1 minimum return inside year one.
- 3:1 = green. Build it.
- 1:1 to 3:1 = amber. Maybe later.
- Under 1:1 = red. Walk away.
Axis 02 · Build
Complexity
How many systems touch the workflow, how stable their APIs are, how much custom glue is required. Higher complexity raises the ROI threshold to match.
- 1-2 systems, stable APIs = green.
- 3-5 systems, mixed APIs = amber.
- 6+ systems, brittle exports = red.
Axis 03 · Stay
Technical stability
What's the chance the underlying systems change in the next 24 months? Vendor lock-in, API deprecation, EMR upgrade cycles, regulatory shifts.
- Stable, owned, documented = green.
- Vendor cycle, predictable = amber.
- Pre-EOL or rumored sunset = red.
The matrix
What the decision looks like.
Read the matrix top to bottom by ROI tier. Same complexity, same stability, different return. The recommendation comes out of the cell.
Complexity →
Simple
1-2 systems
1-2 systems
Mixed
3-5 systems
3-5 systems
Complex
6+ systems
6+ systems
High ROI
5:1+
5:1+
Build nowQuick Win cap. 2-3 weeks.
Build nowStandard implementation. 3-5 weeks.
Build, w/ scope guardPhase 1 only. Bank the win first.
Mid ROI
3:1 to 5:1
3:1 to 5:1
BuildClassic Quick Win territory.
Build if stableConfirm stability axis is green.
Don't buildMaintenance will eat the ROI.
Low ROI
Under 3:1
Under 3:1
Maybe laterStack onto another Quick Win.
Don't buildMath doesn't clear the floor.
Don't buildHard no. Fix the workflow first.
Stability axis is the gate: any red on stability flips the recommendation. We don't build on top of a vendor that's been bought, deprecated, or rumored to sunset.
Worked example
A real audit, scored.
From an active client. Names redacted. Three Quick Wins from the audit, three different recommendations off the matrix.
Predictive alerts, condensers
IoT sensors on rooftop condensers. Surface-area is big, ROI is solid but complexity is mixed across 3 telemetry stacks.
- ROI · 4:1, big maintenance tail
- Complexity · 4 systems
- Stability · vendor consolidating
Full custom field-service app
Owner wanted to replace ServiceTitan entirely. ROI is real but the build is two-quarters and risk is high. We said no.
- ROI · 2.1:1, mostly year 2+
- Complexity · 9 systems
- Stability · greenfield, untested
Apply it to your shop
Want your operation scored?
The friction audit applies this framework to every workflow we surface. You'll leave with a ranked Quick Win list, each one scored on all three axes, with the math in writing.