IBM's multiobjective DSS scored supply-chain reconfiguration on 22 considerations

Reconfiguring a supply chain - shifting a plant, a supplier or a distribution lane - has effects that spread far beyond freight cost into quality, customer responsiveness and strategic flexibility. IBM's managers faced mid-level decisions like these regularly, with no structured way to weigh the financial and nonfinancial dimensions against each other.

A team led by Craig Kirkwood built a decision-support system around a prepackaged multiobjective decision-analysis procedure. Managers and analysts used it in facilitated workshops: 22 considerations were weighted, alternatives were scored, and uncertainty was handled with expert estimates of probabilities. Everything ran in a spreadsheet, so the tool traveled easily between meetings.

The approach was applied to five IBM supply-chain configuration decisions. The results satisfied internal stakeholders that the analysis correctly included financial and nonfinancial considerations along with risk, and produced an audit trail executives could review. IBM came to view the system as a potential template for future supply-chain decisions.

Score supply-chain options on 22 weighted considerations

Configuration decisions are multiobjective by nature, so forcing them into a single cost number is where most analyses lose the room.

Workshop facilitation plus a fixed 22-item frame converts competing departmental intuitions into one comparable scorecard.

Running it in a spreadsheet meant near-zero adoption cost for managers who already lived in Excel.

The explicit audit trail let executive management see why an option won, which is what makes a template repeatable.

Big network decisions stall on money-versus-everything-else debates. A fixed scoring frame that prices quality, service and strategy alongside cost gives executives an audit trail, not an argument.

The paper reports five supply-chain reconfiguration decisions analyzed with the system, stakeholder acceptance of the tradeoffs as fairly represented, and IBM treating the procedure as a candidate template for future decisions. The case is often cited for how multiattribute utility analysis moved from consulting theory into routine corporate use.