A system message is twenty lines of text that most people never see. It sits behind every interaction you have with AI, shaping what the AI pays attention to, what frameworks it uses to structure its thinking, what questions it asks, what it challenges, and what it lets pass. When you use AI without a system message, you are working with an assistant that has been configured by default — helpful, articulate, and entirely ignorant of your management context. When you design a system message for your specific situation, you are building an instrument — a thinking partner calibrated to the disciplines, frameworks, and tensions of your organisation.
The workshop teaches the full craft. You learn how system messages shape AI behaviour by experimenting with deliberately different configurations and observing how the same question produces fundamentally different conversations. You learn the architecture of a system message — role definition, core principles, mode setting, tone guidance, boundary conditions — and why each element matters. You build system messages for your own management context, test them against real management work, and refine them based on what you observe. The process mirrors the management discipline it supports: observe, hypothesise, test, refine.
This is a craft, not a technical exercise. Designing a system message that genuinely understands A3 thinking protocols, hoshin planning methodology, KPI management principles, or the distinction between continuous improvement and business evolution requires deep understanding of both what the AI is capable of and what the management discipline actually demands. A poorly designed system message produces an AI that sounds knowledgeable but reinforces mainstream assumptions — the very assumptions your management architecture is designed to challenge. A well-designed system message produces an AI that thinks within your framework, questions from within your logic, and surfaces what matters in your context rather than what matters generically.
The most powerful application is designing system messages for specific management processes. A system message for Compositional Hoshin extraction that guides an executive through the seven-category observation process, capturing tagged insights, redirecting when the conversation drifts to a different category, and maintaining a running compilation across all categories. A system message for KPI Management that understands the distinction between result indicators and process indicators and challenges any KPI proposal that cannot trace from strategic result to daily controllable signal. A system message for A3 problem-solving that holds the eight-step discipline and refuses to let the user jump from problem statement to countermeasure without proper root cause analysis.
These are not prompts. They are designed instruments, and like any instrument they require calibration, maintenance, and governance. The workshop teaches you to build them. The governance structure — covered in the Process Design Council programme — teaches your organisation to maintain them. Together, they ensure that AI does not become another layer of generic capability but becomes a genuine extension of your management architecture, configured to reinforce the disciplines that make your organisation distinctive.
Course Themes
Course available on Learning Hub March28, 2026
theme
1
Strategy
System messages as Management Architecture
theme
2
Planning
Designing System Messages for Management Disciplines
theme
3
Follow-up
Testing System Messages in Practice
theme
4
Results Analysis
Evaluating System Message Impact
theme
5
Discovery
Advanced System Message Design
theme
6
Capability Development
Building Organisational System Message Capability