Improve tollgate
The Improve tollgate is the project's inflection point: the team transitions from analyzing the problem to implementing the solution at scale. The sponsor's authorization at this gate approves not just the solution's concept, but the evidence that it works and the plan to sustain it.
What must be demonstrated at the Improve tollgate
| Requirement | Evidence | Common gap |
|---|---|---|
| Verified root causes | Statistical evidence (hypothesis tests, regression, DOE) linking Xs to Y | Root causes confirmed qualitatively only — no data |
| Solution selected | Solution selection matrix or multi-criteria decision analysis | Solution chosen based on team preference, not criteria |
| Pilot conducted | Pilot scope, duration, results, and comparison to baseline | Team declares success after two days of pilot |
| Statistical improvement shown | Before/after hypothesis test (2-sample t-test, proportion test, etc.) | Charts show improvement but no significance test |
| Unintended consequences assessed | FMEA on new process; adjacent process impacts reviewed | Fixing one CTQ caused a new defect type downstream |
| Implementation plan | Who does what by when to roll out beyond the pilot | Plan is vague: "will be implemented in Q4" |
| Draft control plan | At least a draft — shows the team has thought about sustainability | Control plan deferred entirely to the Control phase |
The confirmation run: final check before the tollgate
After selecting optimal settings (from DOE or other analysis), the team runs a confirmation experiment at those settings to verify the predicted performance. This is distinct from the pilot: the confirmation run validates the model prediction; the pilot tests the solution in the real production environment over time.
Statistical evidence of improvement
The Improve tollgate requires statistical evidence that the change improved the Y, not just a favorable chart. The appropriate test depends on the data type:
- Continuous Y, one before/after group: Paired t-test (same units measured before and after) or 2-sample t-test (independent samples)
- Continuous Y, multiple groups: One-way ANOVA
- Proportion defective: 2-proportion test (chi-square)
- Process capability shift: Compare Cpk before and after; use confidence intervals on Cpk
The before data is the Measure phase baseline. The after data must come from the same process, same measurement system, under the new conditions — not from a simulation or projection.
FMEA on the improved process
Every new solution creates new risks. Before the Improve tollgate, the FMEA is updated to reflect the new process: new failure modes that didn't exist before, old failure modes that are now prevented, and residual risks that will need monitoring in the control plan. A clean "before" FMEA and a revised "after" FMEA demonstrates diligence.
Risk assessment for full rollout
Moving from pilot to full implementation may affect more processes, more shifts, more locations, or more regulatory requirements than the pilot covered. The Improve tollgate is the moment to identify and plan for:
- Training requirements for all affected operators
- SOP and work instruction updates needed
- Change control or regulatory notification requirements
- Capital or equipment needs not yet secured
- Timing relative to product launches or customer commitments
Try 5 free CSSBB practice questions on Improve phase tollgates and solution validation. No account needed.
Take the free CSSBB diagnosticMore: DMAIC overview · Project charter · VOC & CTQ · SIPOC · MSA in DMAIC · Process capability · Data collection planning · Graphical analysis · Root cause analysis · Multi-vari analysis · Hypothesis testing · Regression · DOE · FMEA · Define tollgate · Change management · Financial benefits · Team dynamics · Improve tollgate · Piloting & scaling · Poka-yoke (BB) · Solution selection · Control plan · SPC in DMAIC · Sustaining gains
CSSBB Quest by Bizarre Act. Original study material; not affiliated with ASQ. CSSBB is a certification of ASQ. Terms · Privacy