Analysis

Why So Many Women Leave Tech Between Mid-Career and Senior Leadership

The pipeline doesn't leak at one point - it leaks at predictable career transitions. What the research shows about why women leave tech in the 5-15 year window, and the interventions that actually change the pattern.

The shape of the problem

In most high-income countries, women now enter technology fields at rates that, while still below parity, are higher than they were a generation ago. The pipeline is bigger than it was. But women leave at higher rates than men do at predictable career stages, and the gap widens through mid-career. By the time you reach senior engineer, staff engineer, and director-level roles, the gender ratio is dramatically more skewed than at entry-level.

This pattern is well documented and substantially studied. The two big takeaways from the research: it's structural rather than individual, and the interventions that work are not the ones that get the most attention.

Where the leakage actually happens

The 3-5 year mark. Women working as software engineers, data scientists, and adjacent technical roles report higher rates of considering leaving the field around 3-5 years in. The reasons cluster around culture, advancement opportunities, and adjacency-creep into non-technical roles.
The first-child transition. Both for women in salaried tech roles and for women in academic research, the first-child transition is statistically associated with downstream career impacts that men don't experience at the same rate. Caregiving infrastructure, parental-leave policies, and re-onboarding practices all matter.
The senior-IC versus management decision. At the 7-10 year mark, women in technical roles are more likely than men to be pushed (sometimes implicitly) toward management tracks even when their interests are in continuing as senior individual contributors. Both tracks have value, but the choice should be the employee's.
The director-to-VP transition. The promotion to vice-president-level roles is where the gender ratio in tech leadership becomes dramatically skewed. The reasons here include sponsorship networks, board-style politics, and outside-hiring patterns.

What doesn't fix it

Several interventions get a lot of attention without producing measurable outcomes:

Single-event unconscious-bias training (without follow-through)
Affinity group activity that's not connected to actual decision-making
Public commitments without accountability for senior leaders
Pipeline programs that don't connect to retention infrastructure

What actually changes the pattern

Structural promotion mechanisms. Promotion processes that rely heavily on self-nomination, manager championing, and political visibility systematically disadvantage women. Promotion processes with structured calibration, peer review, and explicit criteria that are independently applied across managers reduce the gap.
Sponsorship, not just mentorship. Sponsorship - senior leaders actively advocating for specific women for specific opportunities - has been shown to move outcomes in ways mentorship alone does not.
Comprehensive caregiving infrastructure. Companies and institutions with substantial paid leave, flexible scheduling, on-site or subsidized childcare, and re-onboarding support retain women through the first-child transition at higher rates.
Senior-IC tracks that are actually credible. Companies that have built principal-engineer and staff-engineer career tracks with comparable compensation and influence to management tracks retain technical women at higher rates.
Re-entry programs. National re-entry programs (KIRAN, Daphne Jackson, WISET) and institution-level returner programs at major employers have produced measurable outcomes - see our re-entry programs guide.

For individual women navigating this

The honest framing: the pattern is structural, and changing your individual choices can only partially address it. That said, several patterns from observed cases of women who navigated mid-career transitions successfully: