.png?width=207&height=108&name=Mega%20menu%20%E2%80%93%20Featured%20Image%20Template%20(1).png)
UK
France
Germany
Italy
Spain
Poland
.svg)
Australia
Japan
USA
Canada
Most workforce training programs are never independently evaluated on job outcomes. Cyber Advance was — and the results show what accountable workforce development looks like in the AI era.
In June 2026, Mindset, an independent research firm, conducted an endline evaluation of Correlation One's U.S. Department of State–funded cybersecurity training program for women in Jordan and Egypt.
Six to seven months after the program ended, the employment rate among program completers had nearly doubled — from 35% at enrollment to 66% at follow-up — and most of that employment was in the exact roles the program was designed to fill. This page summarizes what the evaluation found, why independent third-party measurement matters, and why the "train and hope" model that dominates traditional education will not survive contact with the AI era.
What did the independent evaluation of Cyber Advance find?
Employment nearly doubled — in the roles the program was built to supply
Cyber Advance for Women in MENA is a 20-week cybersecurity training program implemented by Correlation One with support from the U.S. Department of State. It prepares women in Jordan and Egypt for the CompTIA Security+ certification and for entry into cybersecurity and IT roles. The endline evaluation, prepared by Mindset and administered roughly one year after program entry, surveyed 107 of 160 enrolled participants.
Employment nearly doubled. At baseline, 35% of completers were in paid work. At endline, 66% were employed — and 80% had held paid work at some point since the program began.
Graduates work in the field they trained for. 71% of previously unemployed completers secured work; 71% of new entrants who found work did so as cyber analysts or IT technicians. 62% of all active completers work in the ICT sector.
Earnings rose for those already employed. Median salary increases of 15% in Jordan (675 to 775 JOD/month) and 59% in Egypt (14,500 to 23,000 EGP/month). 85% improved on at least one job-quality dimension; salary gains were statistically significant.
Pass rates were high and consistent. 87% of exam attempters passed CompTIA Security+, with nearly identical rates for those with no prior cyber background (88%) and those with prior experience (86%).
Participants entered the labor market for the first time. Women with no prior employment history now report median monthly incomes of 300 JOD in Jordan and 7,000 EGP in Egypt — entirely new income streams.
Wellbeing improved alongside income. Self-assessed economic standing rose from 5.9 to 6.5 on the Cantril Ladder. Involvement in major household purchase decisions rose from 34% to 52%.
A note on causality: the evaluation is a post-program cross-sectional design, and improvements are consistent with — but not proof of — program impact. That honesty is a feature, not a weakness. Programs that publish independently collected, caveated evidence are the ones funders and governments should trust most.
Why does independent third-party evaluation matter in workforce development?
Self-reported success is the industry default — and it tells funders almost nothing
Most training providers — universities included — report inputs and completions: enrollments, graduation rates, certificates issued. When graduates get jobs, the institution takes credit. When they don't, responsibility quietly shifts to the graduate, the economy, or the labor market. There is no independent measurement, no baseline comparison, and no accountability loop.
Governments, multilateral funders, and philanthropies preparing populations for employment in the AI era cannot allocate capital this way. The World Bank, the U.S. State Department, national ministries of labor, and private co-investors all face the same question: which workforce interventions actually produce jobs and wage gains, and which merely produce credentials? The only credible answer comes from evaluations like this one — baseline-to-endline measurement, conducted by an independent third party, with results published including their limitations.
Cyber Advance's evidence base is exactly what this looks like in practice: a named external evaluator (Mindset), a defined counterfactual discipline (descriptive trajectories, clearly labeled), disaggregated results by country, age, and prior experience, and job-quality metrics — contracts, benefits, satisfaction — rather than placement counts alone.
Why the "train and hope" model fails — and why AI accelerates its decline
Train people, hope they get hired, take credit either way
The traditional higher education model operates on a simple, flawed logic: train people, hope they get hired, take credit when they do, and accept no accountability when they don't. This model is structurally unable to fix itself:
Curriculum is disconnected from employer demand. Course content is set by faculty interest and academic cycles, not by what employers need people to do this year. In fast-moving fields, a curriculum designed on a multi-year committee timeline is outdated before the first cohort graduates.
Incentives reward enrollment, not outcomes. Tuition rises with the availability of student loans, not with the value delivered. An institution gets paid the same whether its graduates find relevant work or not.
The organizational structure cannot respond to labor-market speed. Tenure, departmental governance, and accreditation cycles are designed for stability, not for reorienting a program in months when employer demand shifts.
In the AI era, these weaknesses become fatal. AI is redefining entry-level work faster than any prior technology shift, collapsing the shelf life of static curricula and raising employer expectations for job-ready, tool-fluent talent.
A model that takes four years and six figures to deliver an unverified promise of employability will rapidly lose value against models that deliver measured job outcomes in months.
What does an outcome-accountable workforce model look like?
Not "more training" — a different operating model, orchestrated around current employer demand
Design curriculum with deep employer input
Not advisory boards that meet twice a year — direct involvement from the hiring managers who will interview graduates, shaping what is taught and how competence is demonstrated.
Pre-agree structured job pathways with specific employers
Before a cohort begins, define with named employers what roles graduates will be considered for, what proficiency thresholds unlock interviews, and what the progression looks like.
Use training to advance learners through that pre-agreed pathway
Training becomes the mechanism that moves learners toward committed opportunities, rather than a product sold on hope.
Measure outcomes independently and publish them
Employment rates, wage trajectories, sector alignment, job quality — verified by a third party and reported with baselines and caveats.
This is the model behind Cyber Advance and Correlation One's broader international workforce programs, and the evaluation results show why it works: participants didn't just complete a course — they entered cyber analyst and IT technician roles at scale, in two of the hardest labor markets in the world for women's economic participation.
The evaluation also shows the model's honesty about what remains unsolved. Older jobseekers received far fewer interview invitations than younger ones despite equal search intensity, pointing to employer-side barriers no curriculum can fix alone. New labor-market entrants often start in roles with limited benefits. These are the findings a train-and-hope model would never surface — and exactly the findings an outcome-accountable model needs in order to improve.
What this means for governments and funders preparing for the AI era
The decisive question: does capital flow to models that measure outcomes, or models that count enrollments?
Countries face a compressed window to prepare their populations for AI-era employment. Capital — public and philanthropic — will flow to workforce development at unprecedented scale. Three principles for funders:
- Fund outcomes, not seat time. Tie payment structures to independently verified employment and wage results.
- Require third-party evaluation as a condition of funding. Baseline data, endline follow-up at least six months post-program, and published results with limitations stated.
- Demand employer commitment before training begins. Pre-agreed pathways are the strongest predictor that training converts to employment.
The Cyber Advance evaluation is one data point — but it is the kind of data point the entire field needs to produce as a matter of course. The programs that can survive independent measurement are the ones worth scaling.
Key takeaways
- Employment among Cyber Advance completers nearly doubled — from 35% at baseline to 66% at endline, measured by an independent third-party evaluator six to seven months after the program.
- Placement was aligned with training: 71% of new entrants who found work did so as cyber analysts or IT technicians, and 62% of active completers work in the ICT sector.
- Already-employed participants advanced too: median earnings rose 15% in Jordan and 59% in Egypt, with 85% improving on at least one job-quality dimension.
- Independent evaluation is the accountability mechanism that separates programs producing jobs and wages from programs producing credentials.
- The "train and hope" model is structurally broken — and AI accelerates its decline.
- The alternative is employer-orchestrated: curriculum designed with employers, job pathways pre-agreed with specific employers, training that advances learners through those pathways, and outcomes measured independently.
Frequently asked questions
What is Cyber Advance for Women in MENA?
A 20-week cybersecurity training program implemented by Correlation One with support from the U.S. Department of State, preparing women in Jordan and Egypt for CompTIA Security+ certification and employment in cybersecurity and IT roles.
Who conducted the evaluation?
Mindset, an independent research firm, conducted the endline evaluation on behalf of Correlation One in June 2026, surveying participants six to seven months after program completion.
What were the key employment results?
The employment rate among completers nearly doubled from 35% at baseline to 66% at endline; 80% held paid work at some point after the program began; 71% of previously unemployed completers found work; and 71% of new entrants who found work did so as cyber analysts or IT technicians.
Did earnings improve?
Yes. Among completers employed before the program, median monthly earnings rose 15% in Jordan and 59% in Egypt. First-time labor-market entrants established new income streams with median earnings of 300 JOD (Jordan) and 7,000 EGP (Egypt).
Why does independent evaluation matter for workforce programs?
It replaces self-reported success with verified, baseline-to-endline outcome data, allowing funders and governments to distinguish programs that produce jobs and wage gains from programs that only produce credentials.
How is this different from traditional higher education?
Traditional higher education trains learners and hopes they find work, without accountability for outcomes. Outcome-accountable workforce models design curriculum with employers, pre-agree job pathways before training begins, and submit results to independent measurement.
Build workforce programs that measure what matters.
Correlation One designs employer-orchestrated training programs with independently verified job outcomes — across cybersecurity, data, and AI skills, for governments, multilaterals, and enterprises.