Case study 01 · Delegation and permissions
Inbox Copilot
This case study covers work at Brex and is password protected.
Don't have the password? Email myrto.grigori@gmail.com and I'll send it over.
Context
Every year, hundreds of thousands of ambitious women face the same question: what do I do about work now that I'm a mother? It arrives in the worst possible conditions: physical recovery, a newborn, an identity under reconstruction.
These conditions don't resolve cleanly. The median gap between a first and second child in the US is about 2.5 years, so many women face the decision again before recovering from the first time, with more financial pressure and a toddler in the picture. Each pregnancy is a new chapter in an ongoing story, not a reset.
This isn't hypothetical for me. I am the user, and the absence of a tool like this has already cost me clarity and time.
The data
80%
of mothers report subjective cognitive decline postpartum (JAMA Neurology, 2023)
2 yrs
brain structure remains altered postpartum (Nature Neuroscience, 2017)
2.5 yrs
median gap between first and second child (US Census Bureau)
33%
reduction in female earnings after children (NBER Child Penalty Atlas, 2023)
What women told me
Finding 01
The decision is made at the worst moment.
Most women decide at 6 to 8 weeks postpartum, at the peak of the fog. The tool has to meet them before that and hold their thinking until they're ready.
Finding 02
No tool accounts for the second chapter.
Women on their second leave described starting from zero. The path they chose, what it cost, what they learned: none of it carried over.
Finding 03
The compounding cost is invisible until it isn't.
Women underestimated the cost of two leaves. The surprise wasn't the salary gap but the lost negotiating leverage and network access.
Why existing tools fail
Financial calculators
Compare take-home pay with childcare cost. Miss career trajectory, seniority loss, and re-entry penalty.
HR and benefits portals
Built for compliance, not decisions. Used once per leave, with no memory between them.
Wellness content
Validates feelings without the financial reality. Treats each postpartum period as isolated.
Career coaches
Frameworks that assume continuity. Not fluent in postpartum economics or multiple leaves.
Every tool treats each pregnancy as a new user. The value is in what the tool remembers between decisions.
The journey
Return meets her at five moments per child, from the third trimester, when she is clearest-headed, through the fog peak at 6 to 8 weeks, the end of leave, the 12-month check-in, and the re-entry window. Each moment carries a different cognitive load, so the tool asks different questions at each.
The paths
Full-time
Current role, full hours.
- 100% income retained
- Highest career continuity
- Highest childcare cost
- Leverage strongest at 2nd leave
Reduced hours
Fewer hours or lighter role.
- 65–80% income
- Industry presence kept
- Managing pace, not stepping back
- Most flexible re-entry
Freelance
Own schedule.
- ~75% income potential
- Highest autonomy
- No employer benefits
- Depends on network
Career pause
Step back, return later.
- Zero income
- Maximum presence
- Re-entry modeled at 1, 2, 3 years
- Cumulative cost modeled
New role
New employer on re-entry.
- Salary recovery
- Flexibility negotiated day one
- Search window: on leave
- Fresh culture fit
Any path can be modeled with or without a job change, including what a 15 to 20% salary uplift on re-entry does to the 36-month projection. Paths are hypotheses: Return tracks what was chosen against what actually happened.
The design
A recurring conversation with memory. She describes where she is now, picks up to three paths to model, and sees a 36-month projection of net income after childcare, career trajectory, and wellbeing, with AI insight on demand.
Key interaction
The second pregnancy, when Return already knows her story. This is the interaction no other tool can have.
1
It opens with her history, not a blank form."Last time you chose reduced hours. You described the mental load as manageable by month 8."
2
It names what has changed."Your salary is 12% higher. Your first child starts pre-K at month 6, which changes the cost model."
3
It asks the question she couldn't answer last time."You weren't sure about your employer's flexibility. You now have 12 months of real data. What did you learn?"
Two surfaces
B2C · Direct to mother
- Entry
- Third trimester or early postpartum.
- Value
- Clarity, timing, and memory across the full arc. More useful at chapter two than chapter one.
- Model
- Low monthly subscription, high lifetime value.
B2B · Employer benefit
- Entry
- HR benefit at parental leave notification.
- Value
- Retention. Losing a mid-senior woman after leave is expensive; losing her twice is a talent strategy failure.
- Model
- Per-seat annual license with HR platform integration.
Validation
Return hasn't shipped commercially. Every woman I spoke to recognized the decision immediately, and none had modeled paths side by side. I also ran my own decision through the framework before designing the product: modeling both children, naming the mental load, and having context from my first return produced more clarity in 20 minutes than months of thinking alone.
If it shipped, I'd measure: whether she returns for her second leave (the core signal), whether she uses it in the third trimester or at the fog peak, how closely her 12-month results match the model, and whether employers see lower postpartum attrition.
Reflection
Copilot asked how you delegate without losing control. Price Transparency asked how you act on incomplete information. Return asks both, and adds a third question: how do you design for a decision that repeats, compounds, and gets harder to undo the longer it goes unsupported?
A one-time tool serves a moment. A companion earns a relationship that gets more intelligent over time.