1. Scope the learning
Finished duration matters, but module count and the type of work matter too. An 80-minute course is not automatically equivalent to eight 10-minute modules.
2. Set the production profile
Choose the closest overall experience. The benchmark is a starting point; project-condition adjustments keep it from pretending that duration is the only driver.
Benchmark and advanced production assumptions
Cahill planning heuristics — editable model values
These are illustrative house assumptions, not published industry standards. Edit them to match your environment. A factor of 1.00 makes no adjustment.
| Model component | Value | Applied as |
|---|---|---|
| New build | core factor | |
| Existing-training conversion | core factor | |
| Existing-course refresh | core factor | |
| Approved content | core factor | |
| Structured content | core factor | |
| Rough SME content | core factor | |
| Research/content creation | core factor | |
| Inclusive baseline | core factor | |
| Enhanced WCAG effort | core factor | |
| Regulated/high assurance | core factor | |
| Each review round after 2 | % core | |
| Each active reviewer after 1 | % core | |
| SCORM testing | + | base hrs + hrs/module |
| xAPI/cmi5 | + | base hrs + hrs/module |
| Custom/LTI integration | + | base hrs + hrs/module |
| Added-person efficiency | % capacity per added person |
3. Model an AI-assisted scenario
This is a scenario, not a guaranteed saving. Use a net reduction after prompt preparation, human review, quality control, and rework.
4. Add production costs and capacity
Media and localization remain separate so they are not silently counted twice inside the development ratio.