Skip to content
All Work

Amazon Continuous Improvement at Scale

Building the Amazon portfolio framework, data-backed scorecards, process standards, and operating mechanisms that aligned cross-regional continuous improvement teams and turned ad hoc problem-solving into operating discipline.

Role
Senior Program Manager and Manager, Continuous Improvement
Year
2016–2022
Context
Amazon (2016–2022). European and global customer service operations across multiple sites and regions, with uneven maturity.
Scope
Cross-regional CI operating system · portfolio governance, scorecards, standards, and coaching mechanisms
Result
A governed continuous-improvement operating system across regions: portfolio visibility, shared standards, scorecards, and reusable decision mechanisms.
Continuous ImprovementLean Six SigmaOperating MechanismsProcess DesignGovernance

Slice by domain

See how I apply the same core experience across data, product, program, process, and people leadership.

Summary

Continuous improvement was happening across Amazon customer service, but unevenly. Strong problem-solving in some sites, little in others, and no shared way to see priorities, dependencies, or results across regions. I built the portfolio framework, data-backed scorecards, process standards, and operating mechanisms that turned scattered local effort into a governed, cross-regional discipline.

The baseline

Continuous improvement in a large operation tends to drift toward two failure modes. Either it becomes a set of disconnected local projects with no shared standard, or it becomes a reporting exercise that looks active but changes nothing. Sites were running their own versions of the same work. A good idea in one region had no path to the others. Effort was real. Compounding was not.

Root cause

The problem was not a shortage of Lean tools or trained people. It was the absence of an operating model. There was no common intake for improvement work, no shared definition of what good looked like, and no mechanism to align teams across regions or to make results visible to leadership in a comparable way. Improvement without ownership decays. That was the gap.

The intervention

I designed and governed a portfolio management framework for cross-functional continuous improvement. It gave the work three things it lacked: a common way in, a common standard, and a common view.

A portfolio framework, so initiatives were prioritized and tracked consistently rather than locally. Process standards and standard work, so a method that worked in one site could travel. Operating mechanisms, dashboards, and collaboration workflows that aligned teams across regions and gave leadership a real readout. I ran both Agile and Waterfall delivery depending on what the work needed, rather than forcing one shape onto everything.

Making it stick

The point of standard work is sustainability. A process fix that depends on the person who designed it is not a fix. I built the mechanisms that kept the gain in place: documented standards, operating rhythms, and visibility that made regression obvious before it became a problem. The discipline had to outlast any single project or any single owner.

Outcomes

A governed continuous-improvement portfolio operating across regions, supported by shared scorecards, project standards, and collaboration mechanisms. Data made improvement work more visible, comparable, and easier to scale. Recognition along the way included a WWCS Kaizen of the Year award for the Spain customer service strategy, but the deeper outcome was cultural: improvement became an operating discipline with owners and mechanisms, not a set of one-off wins.

Lesson

Lean fails when ownership is vague. The tools were never the hard part. The hard part was building the operating model that made improvement repeatable, visible, and owned across a large distributed organization. Get that right and the methods take care of themselves.