The part of the overall process which is radically different from most other forms of long-range planning is the central section, the actual production of the scenarios. Even this, though, is relatively simple, at its most basic level. As derived from the approach most commonly used by Shell, it follows six steps:
- Decide drivers for change/assumptions
- Bring drivers together into a viable framework
- Produce 7-9 initial mini-scenarios
- Reduce to 2-3 scenarios
- Draft the scenarios
- Identify the issues arising
NB: There has been next to no research carried out on scenario planning. How does one know that the correct scenarios have been chosen in the first place?