The Mentor Gap for Women Founders Is a Supply Problem

Nothing AVODA has published sorts its mentors or founders by sex. Uganda’s 2021 labor force survey, as compiled by the ILO, counted about 361,000 women employers against 6.6 million women working on their own account: one woman who has hired staff for every 18 who work alone. Among men the ratio was about one to eight. That ratio sets the size of the mentor gap for women founders in East Africa. A useful mentor has already run the business the mentee is trying to build, few women have, and every program draws on those few for free. Women in the Technology and Innovation Ecosystem, a 2026 CRC Press collection on women in American innovation, names the same loop in one short section, though most of its remedies treat the gap as a matter of access and goodwill.

Key Takeaways

In Uganda’s 2021 national labor force survey, as compiled by the ILO, about 361,000 women were employers and about 6.6 million women worked on their own account. That is roughly one woman employer for every 18 self-employed women. For men the ratio was about one to eight.
In the World Bank’s 2025 Enterprise Survey of Uganda, 22.6% of formal firms with five or more workers had a woman as top manager. Kenya’s 2025 figure was 17.7%.
A free mentor hour is drawn on by every program that asks, so the few qualified women carry the whole queue. Asking harder spends the stock faster.
Two tested designs add supply. In Chile, one class visit from a successful graduate raised trainees’ household income a year later, far more cheaply than one-to-one consulting. In Uganda and Kenya, Village Enterprise has run local mentors through groups of about ten businesses.
The Mentor Bench is a four-move plan for a program lead: count the bench, price the hour, share the seat, promote from the cohort.

What the book says about mentors, and the citation behind it

The book’s dominant frame is access. In chapter 8, Rashmi Singh, Anuja Shukla and Sushil Kumar write that mentorship is essential for career development, “but women are often denied these chances.” In chapter 9, Jyoti Suri, Praveen Kumar Pandey and Mohd Asif Shah describe early women leaders who worked “without sponsors or role models who could advocate for them.” The remedies follow the diagnosis: Google’s Women Techmakers, AnitaB.org, the Grace Hopper Celebration and corporate mentoring schemes, each connecting women to senior people who already exist in numbers.

Section 7.12.2, by Harleen Kaur, Rosy Singh, Tilottama Singh and Melissa Hyunh, carries a different idea in one paragraph: senior women in engineering “were less likely to mentor junior women primarily because of their lack of representation.” Few women at the top means few women available to mentor, which means fewer young women entering, which keeps the top thin. The authors call it “an ongoing cycle of invisibility.” That loop is about the size of a stock, and it is the part of the book that travels to Kampala intact.

The citation under it deserves a check. The section credits the claim to Xu (2008), in Research in Higher Education. The paper’s abstract, on the ERIC database, is about faculty turnover: women in STEM were as committed to academia as men, more likely to change institutions, and dissatisfied with research support and advancement. It is silent on who mentors junior women. The supply idea may still hold, but it rests on reasoning more than on the study cited, so the East African numbers have to carry it.

How many women in Uganda have done what the mentee is trying to do

A useful mentor is one stage ahead of the mentee in the same kind of business. For a woman running a market stall alone who wants to hire her first two workers, that means a woman who has hired. The ILOSTAT figures for 2021 give one woman employer for every 18 self-employed women. Among men, 738,000 employers stood against 5.6 million own-account workers. The 2024 census, on its own definitions, gives a similar picture: about 19 to one for women and 11 to one for men.

Higher up, the pool narrows again. The World Bank’s Enterprise Surveys cover formal, registered firms with five or more workers. In Uganda’s 2025 survey, 22.6% of those firms had a woman as top manager. Kenya’s 2025 survey put it at 17.7%, Tanzania’s 2023 survey at 17.5%, and Rwanda’s at 28.1%. A cohort of women aiming at a registered firm with staff is looking for mentors from roughly one formal firm in five.

These are counts of positions, silent on any one woman’s generosity, and many of the women inside them mentor gladly. What they fix is the arithmetic. If every woman employer in Uganda took one mentee, seventeen of every eighteen self-employed women would still be waiting.

Why asking harder empties the bench

Picture the requests from the mentor’s side. A hub runs two cohorts a year, a women’s business association runs one, and a donor program, a bank’s SME desk and a church business fellowship each want mentors. Each writes to the same short list, because the short list is who everyone knows. Each request is reasonable. Together they draw on one shared stock of evenings and goodwill, and a stock everyone uses free gets overgrazed.

Overload has a recognizable shape. While a mentor’s calendar has slack, one more mentee costs little. As requests approach the hours she has, waits lengthen fast, sessions shrink and cancellations arrive. The program concludes that experienced women are unwilling. What it has measured is a queue.

The cost falls hardest on the best-placed mentors. A woman whose business is growing values each evening most, so the women a cohort most needs have the least spare time. Programs fill the gap with whoever is available, and the roster drifts away from people who have run the business. Rating those mentors one by one, the subject of Mentor Quality Control: Rating the Free Resource, helps a founder choose among them. It leaves the pool the same size.

The Mentor Bench

The Mentor Bench has four moves, and each one adds supply or protects it. Together they turn a recruiting habit into a budget line.

1. Count the bench. Write down by name every person within reach who has employed staff in the trades your cohort runs. Start with women, then add the men who have run those trades, since the 2021 survey counted about twice as many male employers. Divide the cohort by the names. If there are more mentees than those people have evenings, the plan is a queue, and the other three moves shorten it. AVODA’s public numbers (600+ trained, 100+ startups supported, 18 programs, three countries) all count founders, so this move applies to us as well.

2. Price the hour. A paid hour is an hour a funder can see in a budget and protect. Village Enterprise Fund described its model to GiveWell in 2009 in these terms: 80 business mentors, each paid a small stipend for transport, grant processing and mentoring, at about 12.5 new businesses per mentor. Mentoring was costed at $22.50 per business for a year, against $9 for the training itself, with $200 per mentor each year for train-the-trainer sessions. Those are 2009 figures from the organization’s own proposal, and the principle carries: the mentoring line was the larger one, and someone budgeted for it.

3. Share the seat. One-to-one matching is the most expensive way to spend a scarce hour. In a Chilean trial of micro-entrepreneur training by Jeanne Lafortune, Julio Riutort and José Tessada, personalized consulting cost almost ten times as much as group assistance. Village Enterprise, as the organization is now called, ran its development impact bond program in Uganda (Soroti, Dokolo and Amuria) and Kenya (Trans-Nzoia and Bungoma) through savings groups of about ten businesses, with local mentors giving four months of training and a year of guidance. IDinsight’s trial, run from 2017 to 2021, found household consumption 6.3% higher than in the control group, and 3.6% in Uganda. That trial measured the whole package of cash, training and mentoring, so it speaks for the group design at scale and leaves the mentor’s own share unmeasured.

4. Promote from the cohort. The cheapest inflow to the bench is your own graduates. In the Chilean trial, where 92% of participants were women, a successful graduate of the program attended one class and shared a testimony. Trainees in groups that got the visit had higher household income a year later, and the authors found the visit significantly more cost-effective than individual consulting and better suited to less experienced businesses. Village Enterprise Fund did the same in 2009, building its expert trainers from inside its own mentor network. A program that asks two graduates from each cohort to teach one session grows its bench every cycle.

Owner-operators mentor on unpaid evenings

The book’s remedies assume a full room behind the door. Women Techmakers, AnitaB.org and corporate schemes connect women in large firms to senior people who are paid by those firms and can mentor in working hours. There, opening the door is the hard part, because the stock exists.

In Uganda the woman founder is usually an owner-operator, and so is the woman one stage ahead of her. Her evening is unpaid, and her business needs her at the hours a cohort meets. The stock is thin, as the employer counts show, so opening the door leads back to the same short list. Imported unchanged, the access remedies raise demand on a supply already at its limit. What does transfer is the loop in section 7.12.2, which runs the same way in Kampala as in an American engineering department. A program that adds graduates to the bench is changing that structure, where one that asks harder only fills a roster.

Overmentored, undersponsored: the Catalyst finding

Herminia Ibarra, Nancy Carter and Christine Silva made the opposing case in the Harvard Business Review in September 2010. A Catalyst survey of over 4,000 high-potential employees found that more women than men had mentors, yet the women were paid $4,600 less in their first post-MBA jobs and held lower-level positions. Their verdict was that high-potential women are “overmentored, undersponsored.” One manager they interviewed put it plainly: “Now I am being mentored to death.” On their evidence, adding mentors only adds sessions.

That should be conceded in full for the setting they studied. Inside large firms mentors were plentiful, and mentors without power changed little. A Kampala program that adds sessions with people who cannot open a door will reproduce their finding.

Their answer, though, is itself about supply. The scarce input they found was the sponsor, a senior person with a seat at the decision who would spend credibility on a junior woman. In Kampala that is the woman who can introduce a buyer or vouch for a younger founder at her SACCO, a capacity that comes only from having run the business. So the objection sharpens the first move: count the people who could sponsor, and expect a shorter list.

Workers for the harvest

In Luke 10 Jesus appoints seventy-two and sends them ahead “two by two,” and he names the shortage they walk into: “The harvest is plentiful, but the workers are few.” He answers it twice, telling his followers to ask for workers and multiplying the ones he has by sending them in pairs. The repair proposed above is mostly budgetary: price the hour, share it, grow the inflow. A funder can act on all three, and payment buys hours. The count of women who have hired staff stays where it was, and a market for her time gives the scarce mentor a reason to sell evenings when the bench needs her to multiply herself. The pairs in Luke 10 carry the reason a budget line lacks. What she learned building her business is held in trust, and her work includes making the next mentors, in pairs, from the people a short way behind her. Programs can recruit for that conviction openly, and ask each mentor to name the two people she is bringing along.

Frequently Asked Questions

Why is it hard to find mentors for women entrepreneurs in Uganda?

Few women have yet run the business a mentee is aiming at. Uganda’s 2021 labor force survey counted about one woman employer for every 18 self-employed women. Every hub, association and donor program asks those few, usually for free, so their time runs out.

Should mentors for women founders be paid?

Paying gives the hour a line in the budget that a funder can see and protect. Village Enterprise Fund paid its 80 business mentors a stipend in 2009 and costed mentoring at $22.50 per business per year, more than its training line. Payment buys hours; growing the number of mentors needs graduates to teach as well.

Is group mentoring as effective as one-to-one mentoring?

In a Chilean trial, one-to-one consulting and a single class visit from a graduate raised income by similar amounts, and one-to-one assistance cost almost ten times as much as group assistance. Village Enterprise runs local mentors through groups of about ten businesses in Uganda and Kenya. Groups let one experienced person reach many founders without spending her whole week.

Can men mentor women founders in East Africa?

Yes, and the counts suggest programs should ask them. Uganda’s 2021 survey counted about 738,000 male employers against 361,000 women. A man who has run the same trade one stage ahead adds to the bench, alongside women graduates who teach.

How can a program grow its own mentors?

Promote from the cohort. In a Chilean trial where 92% of trainees were women, one class visit from a successful graduate raised household income a year later and was more cost-effective than individual consulting. Asking two graduates per cohort to teach one session adds to the bench every cycle.

Next Step

Hold the next mentor recruitment email for a day. Use the day to write the bench down by name, put a mentor rate into next cycle’s budget, and pick the two graduates who will teach. Then book a strategy call with Joshua to test that list and that budget line before the cohort opens. An hour spent counting now saves a term of canceled sessions.

Sources and Evidence

Women in the Technology and Innovation Ecosystem: From Movers to Makers in America (Kaur, Hyunh and Singh, eds., CRC Press, 2026), publisher page: chapter 7 section 7.12.2 (p. 101), chapter 8 (p. 119) and chapter 9 (p. 134), the three quotations and the remedies. DOI 10.1201/9781003604402.
ERIC record EJ804934: Xu, “Gender Disparity in STEM Disciplines: A Study of Faculty Attrition and Turnover Intentions”, Research in Higher Education, 2008: the abstract of the paper the book cites for its mentoring claim.
ILOSTAT, employment by sex and status in employment, Uganda (National Labour Force Survey 2021 and Census 2024): women and men employers and own-account workers, in thousands.
World Bank Enterprise Surveys, firms with a female top manager, Uganda, Kenya, Tanzania and Rwanda: 22.6% Uganda 2025, 17.7% Kenya 2025, 17.5% Tanzania 2023, 28.1% Rwanda 2023.
Enterprise Surveys methodology: survey covers formal, registered firms with five or more workers.
Lafortune, Riutort and Tessada, “Role Models or Individual Consulting”, American Economic Journal: Applied Economics 10(4), 2018: alumni role models raised household income and were more cost-effective than individual consulting.
Lafortune, Riutort and Tessada, working paper version, J-PAL, December 2017: one class visit by a graduate; personalized assistance almost ten times the cost of group assistance; 92% women.
Village Enterprise Fund, Funding Gap Analysis to GiveWell, September 2009: 80 stipended business mentors, 12.5 businesses per mentor, mentoring costed at $22.50 and training at $9 per business, trainers drawn from the mentor network.
Government Outcomes Lab, Village Enterprise Development Impact Bond case study: districts, group size, local mentors, and IDinsight’s trial results.
Village Enterprise, What We Do: group-based, locally led model; over 80% of its entrepreneurs are women.
Ibarra, Carter and Silva, “Why Men Still Get More Promotions Than Women”, HBR reprint R1009F, abstract: Catalyst survey of over 4,000 high potentials; “overmentored, undersponsored”.
Ibarra, Carter and Silva, Harvard Business Review, September 2010: “Now I am being mentored to death.”
Luke 10:2 (NIV) and Luke 10:1 (NIV): the harvest and the workers; the seventy-two sent two by two.