Seeded Expatriate Mentors and Promote-From-Within Ladders Localise Floor Leadership

Passive observation does not transfer complex industrial skills.[S09] Across Ethiopian garment parks, local engineering and management graduates reported that skills transfer occurred mainly through informal job-hopping because foreign supervisors were pressed to meet daily quotas rather than teach.[S09][S1103] GFT prevents operational urgency from crowding out mentorship by enforcing a 1-to-3 Shadowing Protocol synchronized with the four-phase factory ramp (Feasibility Study Section 12) and the four maturity levels of the Performance-Management System (Deliverable 2, Section 7, illustrative schedule).

Each of the eight master mentors is paired with three local trainee deputies in Cohort 1 during Phase 0 (Months -4 to 0) and Phase 1 (Months 1 to 6), yielding 24 certified local floor leaders by Month 6 (illustrative cohort arithmetic). In Month 7, each mentor takes on three local deputies in Cohort 2 (adding 24 local leaders by Month 12, for 48 cumulative graduates), while top Cohort 1 graduates co-mentor 16 to 20 local trainees in Cohort 3 between Months 10 and 16 (illustrative pipeline model). Allowing for an estimated 12 percent vestibule wash-out rate and 10 percent first-year attrition, inducting 86 candidates across the three cohorts secures all 64 certified national leaders before Month 18.[S06][S24]

Authority on the sewing floor transfers in four stages across the illustrative 18-month rollout, as mapped in Exhibit 44:

The schedule in Exhibit 44 staggers handoff so no sewing line experiences an abrupt leadership vacuum (illustrative schedule). Stage 3 and Stage 4 require Cameroonian supervisors and mechanics to run live lines independently for six to nine months while mentors remain reachable on site.[S09] Because this gantt schedule is an illustrative planning baseline, progression from Stage 2 to Stage 3 is gated on competency: if a cohort's lines fail to hold first-pass yield above 93 percent or huddle action closure above 85 percent, the gate holds until the deficit is remediated in the Vestibule Dojo (illustrative gate rule).

A promote-from-within system fails its mathematical purpose if it excludes four-fifths of the talent pool on the sewing floor.[S01] Across global garment hubs, production operators are overwhelmingly women, yet supervision is monopolised by men; in empirical studies of the Bangladeshi garment sector by researchers at the University of Warwick, Innovations for Poverty Action (IPA), and the International Growth Centre (IGC), nearly 80 percent of line workers were female while 95 percent of supervisors were male.[S01][S1101][S1102] In Cameroon, where 93.2 percent of employed women work informally and 86.9 percent experience underemployment, formal garment assembly offers a vital industrial career path for young women.[S42] Yet if a Cameroonian factory copies traditional hub habits, it will recruit women for 80 percent of its Band 1 operator seats while filling 90 percent of its Band 3 supervisory seats with men (illustrative baseline risk).[S01][S42]

Two bodies of randomised controlled trial evidence explain why conventional factory promotions fail and how engineered selection rules fix them:

First, subjective nominations by middle management misallocate training.[S02][S03] In a field evaluation of the STITCH supervisory training programme by economists at the National University of Singapore, lines overseen by treated supervisors raised productivity by 7.3 percent during training and maintained a 5.8 percent productivity edge six months after completion.[S02] Before training, Factory and Floor Managers (FFMs) ranked which supervisors they recommended for training.[S03] Decomposing gains across those rankings revealed an inversion: lines led by highly recommended supervisors showed nearly zero productivity impact, whereas lines led by supervisors in the lowest recommendation tercile achieved productivity gains above 11 percent.[S03] Floor managers direct training toward vocal favourites who have plateaued while overlooking quiet candidates with the largest productivity headroom.[S03]

Second, biased workplace beliefs block capable women from promotion unless firms run structured diagnostic trials.[S01][S1101] In the 60-factory IGC and 24-factory IPA experiments in Bangladesh, objective baseline diagnostics revealed few skill differences between marginal male and female supervisor candidates, and female trainees outperformed men in management simulations.[S1101][S1102] Yet female trainees achieved a promotion rate of only 55 percent compared with 85 percent for male trainees under discretionary promotion.[S1102]

When researchers randomly assigned trained female and male candidates to manage production lines, they found that initial subordinate resistance and negative beliefs caused lower evaluations and a temporary performance dip during a female supervisor's first weeks.[S01][S1101] After four to six months of exposure, however, worker resistance vanished, evaluations equalised, and the productivity of female-led lines caught up completely with male-led lines, leading factories to permanently raise female supervisory retention.[S1101] Moreover, Tufts University evaluations of the ILO and IFC Better Work programme show that lines overseen by Supervisory Skills-trained female supervisors achieve a 22 percent productivity increase over lines led by untrained supervisors.[S20][S23]

The paired comparisons in Exhibit 45 show why GFT removes discretionary manager favouritism from supervisory selection.[S01][S03][S23] While the 5 percent female supervisor baseline, the 55 percent promotion rate, the 11.2 percent low-tercile STITCH gain, and the 22 percent Better Work SST surge are empirical observations, the 60 percent female supervisor share and 82 percent parity promotion rate are GFT's illustrative governance targets for Cameroon.[S01][S03][S23][S1102]

To ensure the four-to-six-month bias adjustment period documented by IPA never derails newly promoted Cameroonian women, GFT enforces three rules:[S1101]