Before you start
On the next page, you will set a resource allocation rule and choose how broadly to apply it. After exploring the case dynamics in the intervening stages, you will return to that choice on The Imprint page and carry it forward over time before revisiting it on the Debrief page.
Two concepts are key to this exercise. Understand them clearly before you begin.
Breadth
How many people get protected capacity to do consequential AI work. Proposal A funds about 45. Proposal B funds about 240.
Depth
How much each one gets. The pot is fixed at $2 million over 18 months, so breadth and depth trade against each other directly.
Where the numbers come from
- 300 employees, 147 women (49.0%)
- Technical core 110, 27 women (24.5%)
- Implementation and customer success 120, 81 women (67.5%)
- Other functions 70, 39 women (55.7%)
- $2 million over 18 months; roughly 45 seats under A, roughly 240 under B
- The 300 individual people. The case reports totals, not individuals, so each person here carries invented attribute scores that let a rule rank them. Every total reproduces the case exactly.
- The five-year projections in The Imprint are illustrative. The parameters were chosen to make the underlying mechanisms visible and interpretable, not to predict Tech Industries' future.
Set Your Allocation
Two separate choices. Make them before you see any results.
Test the Original Proposals
Now test the two proposals from the written case. Adjust how broadly each proposal is applied and notice what changes—and what remains largely fixed.
Sweeps in steps of ten, which is the grid the published band is computed on. At the case's own setting of 45 seats: 24.4% women, $44,400 each.
Case setting: 80%, or 240 people. That gives 49.2% women and $8,300 each.
What just happened
Sweep either slider end to end. Proposal A's women's share stays inside 23.3% to 25.0%. Proposal B's stays inside 48.3% to 49.3%. Neither lever moves composition, because a proportional draw inherits the gender mix of the pool it draws from. Those two shares are restating the source populations, not reporting a result.
Depth is the quantity the levers actually move: from per person at the narrow end of A to at the wide end of B.
Allocation Bench
The breadth you set on the previous page is carried forward to this screen. Now compare five different resource allocation rules. Each rule is applied to the same 300-person workforce, allowing you to see how the method of selection changes where the opportunities go.
| Rule | Depth each | Where the seats land | Women |
|---|
The question to sit with
What did the original A-versus-B framing bundle together that this bench has now separated?
The Imprint
Let's return to the resource allocation rule and breadth you chose earlier. Now carry that allocation forward five years to explore how repeated choices could shape visibility, retention, attraction, and organizational composition.
| Year | Women in the funded pool | Women companywide | Trajectory |
|---|
The question to sit with
If Tech Industries used your allocation approach repeatedly, what kind of expertise would become more visible and influential over time? What kind of company might that create?
Debrief
Let's return once more to the resource allocation rule and breadth you chose at the start. Now consider what the intervening stages revealed about the consequences of that choice.
What you chose
Three questions
- Your resource allocation rule made no reference to gender. Look at the women's share it produced. Where did that number come from, if not from the rule?
- You set breadth and rule separately. The case made you pick one bundle. Which of your two choices was doing more of the work?
- Whoever receives this work becomes the visible example of who succeeds here. What does your allocation teach the next person deciding whether to apply?