Despite decades of promising pilots and innovations that have delivered underwhelming impact, workforce development in the U.S. has nonetheless made genuine progress, argues Matthew Muench, a veteran of major philanthropic pushes on workforce. Muench proposes a four-wave framework to understand the field’s momentum. From the community problem-solvers of the 1980s to today’s AI-native pioneers, each successive generation of workforce organizations has become more commercially sophisticated, better designed for scale, and faster to reach it. The framework is both diagnostic and prescriptive: it helps workforce professionals understand why most organizations remain sub-scale, and reveals what it would take to create a meaningful breakthrough for the field. The proposition is both sober and hopeful.

With generative AI poised to accelerate job churn faster than any previous technological shift, Muench argues the field can no longer afford another decade of admiring promising pilots. The question is whether funders, policymakers, and operators are willing to abandon the habits—and the organizations—built for lesser ambition, and instead create the economic mobility infrastructure of the AI age.

A System to Meet Our Ambitions

The U.S. workforce development system has never had the effectiveness and reach needed to help the mass of workers move quickly into new jobs as the economy changes. And despite decades of promising pilots and talk of scale, the workforce field still operates at a fraction of the scale its own ambitions—and our economic reality—require.

This longstanding problem has new urgency: Generative AI is poised to accelerate the churn of jobs, industries, and skills faster than any technological shift in a generation. Now everyone—educators, policymakers, influencers, billionaires—is proposing solutions to help workers adjust to the coming change and avoid the most dire of the prognostications. The conversations I’ve had over the last year with corporate talent and business executives, philanthropists, academics studying labor markets, and workforce development professionals carry a current of urgency that we need better, faster training solutions to help people adjust to AI’s effects on work and the demand for labor. 

This need is different in degree, not in kind, from what many policymakers and workforce professionals have been working on for a century: helping people transition to skilled, high-wage jobs as occupations and industries decline; addressing talent shortages in critical industries; compensating for the shortcomings of the education system to adequately prepare people for a hard-to-predict labor market; and to increase labor force participation—especially at this moment among working-age men. The task of helping the country keep people productively employed as the wheels of AI occupational destruction speed up is the same work, but our response needs to be bigger, faster, and more effective.

To succeed at this, our new efforts must incorporate the best of what we’ve learned over the last 50 years so that we can finally build the quality and scale we need for an opportunity economy. Absent a renewed intellectual honesty, we will naively recreate what didn’t work in the past and will work even less in future.

But we can find evidence of the possible and insights on how to reach it. I propose a new model for understanding the evolution of workforce development organizations across four successive waves of organizations, each growing more ambitious, more commercial, and more capable of meeting workforce needs at scale. The newest generations might be capable of population-level impact in the aggregate, but only if funders, policymakers, and operators abandon habits built for lesser ambition. We can use these insights to cultivate a workforce ecosystem that generates economic mobility on par with community colleges and four-year universities, which produce 3M degrees a year. 

The Big Goal: Effectively connect 2M-3M people to good jobs—and ensure this country’s response is equal to the challenge.

Today’s Reality

Consider the math. The Harvard Project on the Workforce has identified 17K U.S. workforce training providers offering some 75K programs eligible for the country’s principal public training funding stream. Only 558 of those programs serve more than 50 people. Just 250K Americans are trained each year with those funds, and even fewer placed into quality jobs where they thrive. Ultimately, the research describes a highly fragmented system that doesn’t promote job quality and doesn’t enable scale. And outside of the federally funded system, most programs you’ll find in our cities are successfully connecting people to jobs each year at levels that reach just double or low-triple digits. This is nonsensical from a system-optimization perspective. 

Sober reflection is needed on the gap between our ambitions and our results, and an intellectually honest and creative reimagining of how philanthropy, government, and the organizations themselves should think differently about the path to scaled impact. Creating a network of high-quality and large-scale workforce development providers is an essential if insufficient element of building an opportunity economy for the AI age. 

We need massively more—and more effective—capacity. We need speed and we need scale. But haven’t we been saying that over the last two decades as advanced manufacturing and digital jobs grew, and we worried middle-skill workers and low-income families were being left behind? What have we learned from the last half-century that should give us hope or concern? I find some of both.

Scale, Just Over the Horizon

We have been talking about scaling workforce solutions for decades. One of the most well-regarded labor economists in the country, Larry Katz, discussed in a podcast not long ago how in the early 1990s, while he was at the U.S. Department of Labor, the agency realized most workforce development programs had limited impact, but that some—in particular, sectoral employment programs—were more promising in their impacts, but still small in their reach. The question for them: how to scale these things that were working?

In the 30 intervening years, we have been responding to every labor market shift with a cry for scale to meet demand—through the manufacturing decline and rise, the hollowing out of middle-skills jobs, the rise of digital jobs, and now healthcare and the trades. And the scale-what-works agenda has remained a perennial hit for social sector consultants, workforce development conference panels, government agencies, philanthropies—and, well, authors of social impact articles. Each new iteration of the conversation can feel like a rerun. And annual reports from many of the best-known workforce development groups reveal many years of organizations setting out-year targets to place thousands of workers—goals that were quietly dropped or extended a few years later. 

It’s more a frustrating story than an inspiring one. But if we step back to take a full historical view, we can see progress—we’ve swapped in the right tires and aren’t (always) getting stuck in the same ruts.  

The System We Need

An essential question for anybody trying to build an opportunity economy is what role we want workforce development programs to play in ensuring everybody has access to a living wage job. Follow-up questions include what scale they need to achieve in order to play that role effectively, and whether it is possible to get them to that scale. The answers seem clear:

1) Nonprofit workforce programs should be a viable pathway to good careers for those failed by our educational systems, those who hit a rough patch and need a new start, or those who’ve hit a plateau and need renewed upward momentum, and for whom college degrees aren’t the right fit. Given the anticipated effects of AI on jobs, these categories might soon encompass just about everybody;

2) In aggregate, they should help 2M-3M people annually prepare for, secure and persist in quality jobs—equal to new degree production and about about 15% of the theoretical demand today;

(3) There’s reason for hope, but we should be honest about where we stand in working towards that goal.

The leaders of these organizations have improved so many lives and are rightly celebrated. But the next decade will require more from all of us, including critical analysis of beloved partners, and tough decisions about what we continue doing, and what we sunset. Regional leaders in Minneapolis-St. Paul recently shut down well-liked workforce initiatives when they realized there was no path for reaching the scale that would constitute the solution to a regional problem. We can all learn from that clarity of purpose.  

Four Waves of Progress

A review of some of the best-regarded workforce training organizations over the last 40 years suggests a pattern of improvement with respect to scale, at least partially due to an evolving focus on achieving scale in order to meaningfully address opportunity gaps.

The field’s evolution can be understood through four broad waves of organizations, each defined by its respective era, and each improving on the last in its capacity to scale. 

  • Wave 1: Community Problem-solvers. Nonprofits that emerged in the 1980s and early 1990s (sometimes earlier) to solve domestic employment challenges resulting from deindustrialization, urban poverty, and neoliberal economic policy.
  • Wave 2: Social Entrepreneurs. Millenium-era leaders alongside the broader phenomenon of social entrepreneurs taking a more business-like approach to operating, fundraising, and measuring and articulating return on investment in the era after welfare reforms. 
  • Wave 3: Founders. Leaders after the financial crisis aped the aesthetic, approaches, and tools of the venture-backed startup culture and sought rapid growth right out of the gate.
  • Wave 4: AI-first pioneers. Emergent post-ChatGPT providers are attempting to leverage the capability of generative AI tools to resolve the scale vs. personalization conundrum that has foiled education and social-services entrepreneurs.

Wave 1: The Community Problem Solvers

An initial wave of workforce development organizations using modern models emerged in the 1980s and early 1990s and established a reputation for impactful work but haven’t grown much beyond their initial reach.

Proto-social-enterpreneurs around the country responded to the unemployment and poverty challenges created by deindustrialization—and then the support gaps left by welfare reform—and jumped into the fray to help people find jobs and train for better ones. The founders were generally educators, community organizers, and social-work practitioners who saw existing inadequacies in our systems and stepped up to fill gaps. Several of these pioneering organizations reside in the pantheon of storied workforce development programs and are bywords for effective workforce development. See Project QUEST in Texas, STRIVE in New York, TowardsEmployment in Cleveland, TwinCitiesRise! In Minneapolis, and i.c.stars in Chicago (also notice the geographic dispersion). These groups continue to show up in books on how to address opportunity gaps, help people out of poverty, or create dynamic opportunity economies. And they play an important role in the workforce-development believer’s psyche in demonstrating that it is possible to do what we’re trying to do—the gold-standard randomized controlled trials, particularly of QUEST and TowardsEmployment, verify that high-quality programs improve living standards by increasing employment and wages. 

These organizations are doing very impactful work for many individuals and families each year—we celebrate this. Yet, their psychological impact on the workforce field is perhaps disproportionate to their quantitative impact. Of the many of these programs that were launched, the two or three largest grew to place around 200-300 people in jobs per year by 1995, but still place only around 500-1K people annually 30 years later. The largest of them today, STRIVE, places more than 1,500 in jobs across their multiple STRIVE-operated sites, with credible plans to grow further. They are the laudable exception. As a class, with modest scale after more than three decades, these organizations remain guiding lights but seem unlikely to change the opportunity dynamic of cities.

Wave 2: The Social Entrepreneurs

The late 1990s saw a generational shift in the social sector, following the passage of welfare reform, complementing the “third-way” politics of Bill Clinton and Tony Blair, and incorporating business thinking and practices into social work. The idea of social entrepreneurs and social enterprises came into vogue. REDF was founded in 1997 to support social enterprises that created employment opportunities. The field gained professional credibility with the launch of the “Stanford Social Innovation Review” in 2003, then global validation with Mohammad Yunus winning the Nobel Peace Prize in 2006 for pioneering the field of microfinance. David Bornstein’s 2007 book, “How to Change the World,” inspired a generation of young professionals seeking to make a big difference and not finding it in the drab offices of traditional, slow-moving nonprofits.

Social-entrepreneur (though not necessarily social enterprise) workforce development organizations were a part of this story. These new workforce organizations sought to balance mission with “operating like a business,” striving for excellence, using data and modern management tools to demonstrate impact, enable operational improvement, and fundraise more aggressively. Whereas Wave 1 leaders were community organizers or nonprofit leaders, Wave 2 founders were managerial scalers—they often had business backgrounds and saw opportunities to leverage more effective business management approaches to address employment challenges at larger scale or for more valuable niches, like tech. The ones that gained significant traction were mostly founded in New York or Boston. They more loudly told the story of their impact, adopted a national rather than local vision, and gradually found growth via establishing new site operations in cities beyond their original foundation. This is the era of nonprofits starting to release impact-focused annual reports mimicking the approach of Fortune 500 companies. 

The result of this is demonstrable: whereas I find only one Wave-1 organization exceeding 1K annual job placements today, Wave 2 has several. YearUp is the standard-bearer for this cohort with nearly 5K annual placements. Per Scholas, NPower, and the Center for Employment Opportunities also have validated impactful models and place more than 2,500 people per year into jobs. The shift from Wave 1 to Wave 2 is significant—you can see it gradually but clearly emerge in the chart. And several of them complemented their growth with participation in RCT evaluations to validate their impact. 

Again, we celebrate the impact on tens of thousands of families, and most workforce providers can learn a lot from the Wave 2 leaders about delivering impact and growth. But it’s difficult to argue this is a meaningful shift from a population-level impact perspective. Only a few of these organizations effectively serve more than 1K people. And the impact is diffuse—most of the growth in topline numbers comes from adding more sites, with YearUp, Per Scholas, and CEO each in more than 20 cities (and most have some virtual component). That translates to 100-200 people placed in jobs in some of our largest cities—again, changing individual lives but not communities or cities.  

Fortunately, the story of the leaders of this class is still being written, and they continue to innovate their models and pursue growth aggressively—in many ways coming to look more like the Wave 3 organizations. I write this article as a commentator, but I’m also involved in the work. And I’ve been impressed by my colleagues running these organizations; they see their current footprint as the starting point for the next phase of growth and impact, and we should all root for them to keep going and figure out how to help them get there. And in addition to these Wave 2 leaders continuing their upward climb, we need many more to follow their path into the 1K+ category if we’re going to reach the goal of a 2M+ workforce system.

Wave 3: The Founders

Like the idea that a new garden sleeps, then creeps, then leaps, we see big blooms early in the season from Wave 3. This most recent group of providers has more of a startÏup orientation, designing their model for scale from the outset and building the personnel and organizational capabilities to fundraise, sell, and deliver customer success. The leading groups in this category—Merit America, Generation, Apprenti, etc.—have achieved 1K+ people served annually within several years of launching, which is much faster than earlier waves. And they appear to be pursuing bold approaches to expansion, such as with Merit pursuing accreditation federal financial aid eligibility, and Apprenti pursuing both M&A strategies and raising a novel fund for apprenticeship expansion.

If Wave 2 was a Clinton-era phenomenon, Wave 3 feels like the Obama generation. These groups are largely tech-positive, developed during a time when “lean startup,” “blitzscaling,” and “agile” gained traction with social entrepreneurs. Interestingly, “social entrepreneur” plateaued in usage in 2018 as these new workforce groups emerged. That moniker had started to feel dated and lacking the rapid-scaling aggression and sophistication of venture-backed startups these companies learned from. Looking across the most successful of these organizations so far, you’ll see several founders and senior executives with prior experience at built-for-scale software companies as well as elite consulting firms oriented ruthlessly around strategies to win. This is a contrast to the Wave 1 and Wave 2 founders—even those in the earlier waves who came from business did not come from software, with the exception of the founder of the Wave-2 leader YearUp United.

Certainly, scale challenges for all of these groups are legion. Even the biggest companies that are committed to hiring from nontraditional channels like these tend to source paltry numbers from any individual organization today—especially amid the current slow-down in entry-level hiring. But these organizations have demonstrated their ability to quickly grow to effectively serve many people, and are actively pursuing additional scale. To my mind, this is good and we need more of it. 

Wave 4: AI-first Pioneers

We are still relatively early in the story of the Wave 3 organizations. But generative AI could well upend everything and shift us quickly to a Wave 4 that is truly AI-first. The winners of the current crop of new organizations will leverage generative AI for scale the way existing orgs leverage SaaS tools. And as is being daily debated right now, when AI is more widely incorporated into productive use in companies, it will reshape the entry-level labor market and may require entirely different types of workforce development solutions. Let us hope those who step up to provide those services also ensure they are learning the lessons of Wave 1, 2 and 3, and are able to provide a significant bulwark against the harmful effects of task automation and cognitive replacement, providing truly valuable economic mobility support to many thousands of people. Already, entrepreneurs are showing how AI can help break through barriers that have held back tech-enabled career-navigation and coaching tools; accelerate technical-skills development through more personalized learning with unlimited reps and targeted feedback; and narrow the capability gap in cognitive roles for those with lower levels of education. Can AI help break the trade-off between scale and serving populations with multiple barriers that we saw in the first three waves?

Counterpoints

Several caveats are needed here. First, large organizations such as Goodwill Industries International, YouthBuild, and community colleges are all important, but their business models, funding sources, and goals are different in kind than what we’re discussing.

Second, I acknowledge the trade-off between scale potential and the population an organization serves—the more barriers to employment faced by the population an organization chooses to serve (e.g., homelessness, criminal convictions, addiction), the more difficult it is to scale with effectiveness. Nearly all of the groups discussed here focus on low-income individuals needing good jobs, and most emphasize those without college degrees. But some—think Roca in Boston or Cara in Chicago—intentionally serve the homeless, or formerly incarcerated, or those who didn’t finish high school. That is more difficult work, requiring more intensive human support, and also produces jobseekers who generally are less competitive for high-paying jobs than somebody in stable housing with a work history and some time at a community college. This approach by nature is more difficult to scale. And a focus on these populations is more common in Wave 1 organizations than Wave 2, and more common in Wave 2 organizations than Wave 3. This shouldn’t lead us to dismiss what the later waves have to teach us about scale. It is noteworthy that the Center for Employment Opportunities serves this multiple-barriered population and still has managed to grow its impact to serve thousands per year.

Third, this analysis focuses on the organizations themselves, without acknowledging the problematic funding environment they navigate, including fragmented and onerous public funding; short-term grant cycles; underinvestment in growth capacity; funding priority for activity not outcomes; local incentives counter to scale ambitions. Those are all things to work on, but organizations need to find their path to growing impact in the meantime.

Fourth, there is a line of analysis worth pursuing that flows with the canonical texts of workforce development from the last 30 years—namely, that scaled impact can’t be achieved by individual organizations growing their reach, but can only be achieved through systems-change, intermediaries, alignment with economic development policy, etc. The two paths to large-scale impact are not mutually exclusive. My strong belief is that we need to help our most successful workforce providers expand their impact as individual organizations. Changing public systems and business practices are powerful complementary levers; they don’t obviate the need for high-performance workforce organizations. 

Conclusion 

The workforce field must avoid another decade of admiring promising pilots and modest growth. If AI disrupts labor markets faster than previous shocks, rapid scale with quality is essential. The 4-Wave Model proposed here shows we *are* making progress, if not quite fast enough. We can use the insights here to inform our collective efforts—where we invest, how we build organizations, what we advocate for, what we prioritize—to get us closer to that goal over the next decade.

Workforce organizations can borrow several insights from this analysis to reach greater scale:

  • Explicitly prioritize growth. More than anything else, an orientation to scale is correlated with delivering impact at scale. As each successive wave of workforce organizations prioritized reaching greater numbers of people, they have done so.
  • Design for scale. Building the organization and the programmatic model for the true meaning of scale, by serving more people and delivering the same (or better) outcomes without a proportional increase in cost and staff. Early-wave organizations sought growth through replication, or attempted to retrofit their programs for scale. But later founders designed their models that way from the start. Certainly, this involves trade-offs between who you serve, how you serve them, and what value you deliver for them—not every organization can be serving 10K people. But every organization can make more or less scale-enabling decisions, and leverage new technology to do so.
  • Use AI to reduce costs and enhance services. Organizations starting in 2026 are unlikely to need this advice, but the existing landscape of workforce organizations can almost universally better use technology for improving quality of outcomes and number of people effectively served. Every organization should be thoughtfully using new tools to reduce the administrative burden on staff and improve their effectiveness with client engagement, while offering more personalized, constantly available, and effective tools for learners/jobseekers. 
  • Treat employers as customers on whom your survival depends. Workforce development organizations are ultimately B2B firms selling talent to employers. Operating with such a commercial orientation enables success. Offering a solution to a problem, implementing professional sales methodologies, effective salesforce design, sales training, and sales performance management will help increase performance on what is ultimately the most important thing these organizations do—get somebody a job where they can thrive.
  • Be aggressive. In selling and in building partnerships, be unapologetic and dogged. In fundraising, bring bold asks with compelling plans for how to effectively deploy $10M, $50M, $100M. I’ve worked for three different philanthropies and within a large corporation and I’ve seen and felt the difference between organizations that are aggressive and those that are a bit more genteel. The former are more likely to close the deal, get big dollars, and secure customers.
  • Adapt as markets change. The economy, labor market, funding environment, policy context, and competitive landscape will change. Change with it. Looking at the organizations on our list that have achieved the biggest reach, they have all continued to try new sales channels and partnership models, make acquisitions, secure new legal status, raise new forms of capital to fund their work, etc. They aren’t sitting comfortably on their success, but asking how to keep growing and expanding their impact. As a result, they’re having a lot of it.  

Here’s how we can shift our investment philosophy to enable this new orientation: 

  • Policymakers. Consider the funding and program-design levers you can pull to focus grantee efforts on what will lead to effective scale, incentivize scale-oriented investments by workforce organizations, and reward progress in growing the number of people they’re placing into quality jobs.
  • Philanthropy. Fund growth capacity of the organizations with the model and leadership ready for scale. Don’t simply fund service delivery. For direct-service funders, use the characteristics outlined above as screening criteria, or fund external support for grantees to improve on one or more characteristics—business model advising, technology consulting and money for software licenses, etc. If you fund policy, ask how federal or state workforce policy could be amended to incentivize or support adoption of these practices, and to reward success in serving more people. That is, support a shift to funding for outcomes or impact with a high bar. 
  • Businesses. Partner with the organizations best able to solve talent needs with a commercial orientation—it’s more likely to deliver what you need and to expand opportunity at large scale, and you’re more likely to stick with it.

The Bottom Line: The shortcomings of our workforce system have for decades been an enormous disservice to those struggling towards economic security. They also are a betrayal of our national compromise to intervene where flexible economic policies create dislocation. The failures have not constituted a national crisis because the job-creation success of the U.S. economy has largely compensated. But employment shocks from technology and new business models mean we no longer have that backstop.

We need to get much, much better, very, very quickly. That will require a willingness to put into practice the insights from the last 40 years, accept that some of our beloved partners won’t deliver the volume of opportunity we need, and turn our attention to building a system that can connect at least 2M Americans to good jobs and a career path. It will be enormously difficult. But it is achievable, and it is necessary. 

Matthew Muench has held workforce leadership roles across philanthropy, government, business, and nonprofits for 15 years. He recently founded Lucerra Impact Advisory after serving as Head of Jobs & Skills for JPMorganChase Global Philanthropy.