Manifesto
Improvement starts with real work
The philosophical core of the M.A.G. method (Measure, Automate, Grow): seven convictions and nine principles for improving processes with people, rules, automation and AI, in a verifiable way.
Seven convictions
We believe improvement begins with understanding real work.
Every process must have a recognizable purpose, clear responsibilities and observable results.
Technology finds its place when it contributes to that purpose.
People, rules, automation and AI work together, with the degree of autonomy that suits each activity.
Important decisions must be reconstructable, and exceptions must have a path.
Expected value guides the investment; verified value guides the next step.
Growing means using resources better, serving customers better and making improvement repeatable.
Nine principles
Nine commitments the method does not negotiate
What we believe becomes what we do, in every cycle, with every customer.
Start from the customer's outcome
Define which need is being met and what makes an outcome correct. A process can end correctly even by rejecting an unsuitable request.
Make real work visible
Represent steps, waiting, handoffs between departments, tools, documents and exceptions; compare the declared procedure with actual cases.
Measure in proportion to the decision
A small experiment can start from declared estimates; a significant investment requires stronger evidence. Data collection must be sustainable for operators.
Simplify before automating
Remove pointless activities, clarify responsibilities and stabilize the necessary steps. An organizational simplification can be the best intervention.
Assign each activity to the right performer
Rules and code for deterministic tasks, integrations for structured exchanges, AI for interpretation and ambiguity, people for judgment and accountability. The choice weighs quality, cost and maintainability together.
Govern across the whole cycle
Establish access, required data, limits, approvals, error handling and responsibilities from the measurement stage. Governance runs through Measure, Automate and Grow.
Design for daily work
Test real cases, missing data, errors, human intervention and restarts. Going live includes instructions, adoption and taking ownership of the service.
Make value verifiable
Keep baselines, hypotheses, costs, changes and results. Distinguish freed time from actual spending reduction and from expected economic value.
Consolidate and repeat
Each cycle produces reusable knowledge, a decision on the results and the next improvement. The customer takes part in verification and keeps an understanding of their own process.
Independent of the platform
The principles remain understandable to anyone who has yet to choose a platform. FlowSharp is the reference implementation of the method because it keeps process, execution and evidence connected: see the dedicated page.