Scenario planning

A single forecast invites a single plan, and a single plan breaks when the future turns out differently. Scenario planning prepares for several plausible futures instead.

A planning team might build three to five scenarios for next year:

• Base: demand grows 8%, costs stable.
• Boom: a competitor exits, demand grows 20%.
• Squeeze: a slowdown cuts demand 5% while raw material prices rise 15%.
• Disruption: a key port or supplier is out for six weeks.

Each scenario runs through the plan: what would capacity, inventory, sourcing, cash and service look like? The goal isn’t to predict which one happens. It’s to find:

• robust decisions that work reasonably well in every scenario, such as qualifying a second supplier or keeping a flexible co-packer contract,
• contingency plans for specific scenarios, ready to use, such as a pre-agreed overtime plan or alternative routes,
• trigger points that tell you which future is unfolding, like “if orders exceed plan by 10% for three weeks, start the boom playbook.”

The approach was made famous by Shell in the 1970s, whose scenario work helped it respond to the oil shocks faster than many competitors. In supply chain planning, IBP processes increasingly use scenarios to discuss risks and options with leadership, and digital twins make it faster to test them.

From uncertainties to scenarios

1. List the driving forces that could shape the next few years: demand, input costs, competitors, regulation, technology, geopolitics.
2. Rank them by impact and uncertainty. High-impact forces that are fairly predictable (for example, population growth) go into every scenario. The two most impactful and most uncertain become the critical uncertainties.
3. Cross them in a 2 × 2 grid to get four distinct futures.
4. Write a short story for each and give it a memorable name.
5. Quantify: demand by region, costs, lead times and capacity in each scenario.

Sorting the actions

Run the plan through each scenario, then sort possible actions:

• Robust moves, good in all four: qualifying a second supplier for the main raw material, a flexible co-packer contract, reducing changeover times.
• Contingent moves, right only in some: building a new line (good if demand grows fast), long fixed-price contracts (good if costs are volatile).
• Signposts: measurable indicators that show which future is unfolding, with thresholds. For example: three months of orders above plan by 10% triggers the expansion plan; a raw material index up 20% in a quarter triggers the hedging and pricing playbook.

Do the robust moves now, prepare the contingent ones, and watch the signposts.

What Shell learned

Pierre Wack, who led Shell’s scenario work, wrote in 1985 that scenarios are valuable only when they change how managers see the world. Early scenarios that merely listed possible numbers had little effect. Shell’s breakthrough came when scenarios described the forces that could break assumptions managers took for granted, such as the stability of oil supply, so leaders recognized the signals when the oil shocks came and acted faster than competitors.

Base, plus and minus is not scenario planning

Adding 10% and subtracting 10% from a forecast tests sensitivity, which is useful, but it keeps the same story with different volumes. Real scenarios differ in their logic: different customers, costs, competitors or constraints. They should feel uncomfortable, and at least one should challenge the company’s favourite assumption.

Pre-mortems and reverse stress tests

• Pre-mortem: imagine it is a year from now and the plan has failed. Each person writes down why. The exercise, described by Klein (2007), brings out risks people hesitate to raise in normal planning.
• Reverse stress test: start from an unacceptable outcome (losing the largest customer, missing a month of supply) and work out which combinations of events would cause it.

Both feed scenarios with the risks that matter most.

Linking scenarios to planning

• IBP and S&OP: scenarios become part of the monthly or quarterly review, with the signposts tracked as standing items.
• Capital decisions: investment cases show performance in each scenario, not one forecast.
• Digital twins: network models quickly quantify each scenario’s effect on cost, service and inventory.
• Refresh: scenarios are revisited when signposts move or the world changes.

Common pitfalls

• Too many scenarios, so none gets attention.
• Arguing about probabilities instead of preparing responses.
• Scenarios written by a central team and ignored by the business.
• No signposts, so nobody knows when to act.