There is a category of company that is harder to underwrite than a conventional SaaS business, harder to explain to a committee than a marketplace with proven unit economics, and harder to value than a platform with a clear comparable set — but that has the potential to generate returns that are structurally unavailable to companies operating within existing rules rather than rewriting them. These are the companies that do not simply build better products. They change the ecosystem in which their products operate. They shift standards, restructure incentives, create new categories of participation, and leave their markets fundamentally different from how they found them.
We call this class of opportunity systems change investing. It is not a new concept — social entrepreneurs have used the language of systems change for decades, and the impact investing community has developed extensive frameworks for thinking about interventions at the level of market structure rather than individual product performance. What is newer, and what we find increasingly compelling as a lens for technology investing, is the application of this framework to companies that are systems-change-scale in their ambition but venture-scalable in their economics. These two properties have historically been difficult to combine. The AI era is making their combination more common.
What Systems Change Actually Means
Before applying the concept to investment decisions, it is worth being precise about what systems change means and what it does not mean. Systems change is not simply disruption in the Christensen sense — the process by which a cheaper, simpler product enters from below and progressively takes market share from incumbent providers. Disruption changes who captures value within an existing system. Systems change restructures the system itself: it changes the rules by which value is created and distributed, the identities of the participants who can participate in value creation, and the metrics by which success is measured.
The clearest historical examples come from sectors where the system being changed involved significant structural barriers to participation. Education is one such sector. The traditional K-12 and higher education system in the United States is structured around physical proximity (students must attend school in person), temporal synchrony (students must attend class at scheduled times), and credentialing gatekeeping (value is delivered through institutions with the authority to grant recognized credentials). Each of these structural features creates barriers to participation and efficiency losses that a systems change approach can address.
Khan Academy's intervention in this system is instructive. By making high-quality educational content freely available on the internet, organized according to a mastery-based learning model rather than a time-based model, Khan Academy did not simply offer an alternative to traditional education. It demonstrated that the assumptions underlying traditional education — that quality instruction requires physical presence, that pace must be synchronized across learners, that access requires institutional affiliation — were artifacts of a particular historical technology environment rather than fundamental necessities. That demonstration changed what became possible for everyone operating in the education ecosystem: it shifted what policymakers, parents, and learners considered feasible, and it created the conditions for a broader transformation of how educational value is created and distributed.
The Commercial Systems Change Model: Duolingo and Coursera
Khan Academy is a nonprofit, which raises an obvious question: can systems change investing be commercially sustainable? The answer, demonstrated clearly by companies like Duolingo and Coursera, is yes — but with specific structural requirements that distinguish commercial systems change companies from both conventional SaaS businesses and social enterprises.
Duolingo's trajectory is a case study in building a commercially successful systems change company. The language learning market before Duolingo was dominated by expensive, time-intensive programs — Rosetta Stone's boxed software sold for $200–$500, and in-person language instruction cost hundreds to thousands of dollars per course. The market was structured around the assumption that language learning required significant financial investment and was therefore primarily available to affluent consumers.
Duolingo's core systems change insight was that language learning could be made free and effective through a combination of gamification, spaced repetition algorithms, and a freemium business model that funded free access through premium subscriptions and, subsequently, advertising. The company's 2021 IPO at a valuation of approximately $5.1 billion reflected the scale it had achieved: over 500 million registered users across more than 40 languages, with the structural proof of concept that a free, high-quality language learning experience could be built and sustained commercially.
The systems change impact was not just the free tier — it was the demonstration that language learning could be made genuinely accessible at population scale. By shifting the market expectation from "language learning costs hundreds of dollars" to "language learning can be free and effective," Duolingo restructured the competitive landscape for every player in the language learning market. It also created a data asset — hundreds of millions of learner interactions — that compounded into an AI-training advantage that further improved the product quality and reinforced the competitive position.
Coursera represents a parallel trajectory in higher education. The company, which went public in 2021, built a platform that enables universities to distribute educational content at internet scale, eliminating the physical attendance and institutional proximity requirements of traditional higher education. Coursera's partnerships with more than 300 universities and companies create a credentials marketplace that is recognized by employers globally — slowly shifting the credentials ecosystem toward recognizing online learning credentials that previously carried no formal recognition.
What Coursera accomplished at the ecosystem level is the harder-to-measure but more significant part of its value creation story: by legitimizing online credentials from recognized institutions, the company began to shift employer hiring norms in ways that compound over time. As more employers recognize Coursera credentials, more learners invest in them, which increases the value of the credential ecosystem, which attracts more institutional partners, which improves credential quality and recognition. The flywheel is not simply a product engagement loop — it is an ecosystem legitimacy loop that operates at the level of labor market norms.
The Hugging Face Model: Open Infrastructure as Systems Change
Hugging Face represents a different type of systems change company — one that has used open-source infrastructure as the primary mechanism of ecosystem transformation. The company, which raised $235 million in a Series D at a $4.5 billion valuation, built the dominant model repository and collaboration platform for the AI community. With more than 500,000 models and 100,000 datasets available on its platform, Hugging Face has become the GitHub of AI — the default infrastructure layer through which the global AI research and development community shares, accesses, and builds on AI capabilities.
The systems change impact of Hugging Face's open model is profound and often underappreciated. By providing free access to pre-trained models and the tooling to fine-tune and deploy them, Hugging Face dramatically lowered the barrier to entry for AI development. A startup that would previously have needed to spend millions of dollars training a foundation model from scratch can now fine-tune a Hugging Face model for a specific domain application at a fraction of the cost. A researcher at a university in a developing country without access to major cloud computing budgets can access state-of-the-art models through the Hugging Face Hub.
This democratization of access has restructured the AI ecosystem in a specific way: it has shifted the locus of competitive advantage from model training capability (which requires massive capital and compute resources) to domain application and fine-tuning capability (which requires deep domain knowledge and data, not just capital). This shift has enormous implications for the distribution of AI value creation globally and for the types of organizations that can participate competitively in the AI economy.
The commercial sustainability of Hugging Face's model rests on the insight that open infrastructure creates an ecosystem of users that can be monetized through enterprise services. The free, open-source tier creates a vast user base and a powerful community that generates network effects and improves the platform. The enterprise tier — managed inference, private model hosting, enterprise security and compliance — captures the commercial value created by organizations that need production-grade infrastructure for their AI deployments. This is the same flywheel that has powered the commercial open-source software businesses of the past two decades, applied to AI infrastructure.
The Investment Characteristics of Systems Change Companies
Having examined several concrete examples, it is worth characterizing the investment-relevant properties of systems change companies more precisely. These properties help explain both why systems change companies are difficult to underwrite using conventional metrics and why they generate returns that are structurally unavailable in conventional businesses.
Non-linear scaling. Systems change companies frequently exhibit long periods of apparent under-performance followed by rapid, non-linear growth as the ecosystem they are building reaches critical mass. Duolingo grew slowly for years while it established the credibility and user base required to shift market norms around language learning. The platform effect then accelerated adoption in ways that conventional extrapolation from early growth rates would not have predicted. Investors using conventional growth rate analysis — projecting forward from current revenue or user growth — systematically undervalue companies in the critical-mass-building phase of their development.
Ecosystem value versus product value. The value created by a systems change company is not fully captured in its product metrics. The ecosystem restructuring that Khan Academy accomplished — shifting the market expectation about what quality education should cost — has economic value that does not appear in Khan Academy's own financial statements. Similarly, Hugging Face's contribution to the broader acceleration of AI deployment creates economic value that appears in the performance of thousands of other companies building on its infrastructure, not just in Hugging Face's own revenue. This ecosystem value is real but difficult to capture through conventional valuation approaches, which contributes to systematic undervaluation of systems change companies at growth stage.
Mission alignment as a moat. Systems change companies that have genuine mission alignment — where the founder's identity and the company's product are organized around a clear statement about how the world should be different — tend to build user communities and cultural capital that create durable competitive advantages. The Duolingo community, organized around the goal of making language learning accessible to everyone, is not simply a user base. It is a constituency that has an interest in the company's success independent of the product's features in any given moment. This constituency-as-moat is difficult to replicate through product investment alone and is one of the most durable competitive advantages available to systems change companies.
Regulatory tailwinds. Companies that are aligned with social goals — expanding access to education, democratizing AI capabilities, reducing friction in expert services markets — tend to face favorable regulatory environments rather than adversarial ones. This is not guaranteed, and systems change companies can attract regulatory attention when they restructure markets in ways that affect politically powerful incumbents. But the general principle holds: a company whose success is aligned with broad social benefit has structural advantages in the regulatory environment that a purely profit-maximizing company lacks.
The Diligence Challenge: Evaluating Systems Change Potential
The most significant challenge in systems change investing is the difficulty of evaluating, at early stage, whether a company has genuine systems change potential or is simply a well-intentioned but limited product. The signals of systems change potential are different from the signals of conventional product-market fit, and investors who evaluate systems change companies using only conventional metrics will systematically pass on the most interesting opportunities.
The signals we look for are organized around ecosystem impact rather than product metrics. How are other participants in the ecosystem responding to the company's presence? Are incumbent providers changing their behavior — pricing, product development, hiring — in response to what the company is doing? Are policymakers, academics, or industry associations engaging with the company as a participant in the ecosystem conversation rather than simply as a software vendor? Is the company building a community or constituency around its mission, rather than simply accumulating users?
We also examine the founder's theory of change: their articulation of the mechanism by which their company will change the ecosystem, not just capture market share within it. Founders with genuine systems change ambition can typically articulate this theory of change with specificity — not "we will disrupt the education market" but "by demonstrating that mastery-based learning at scale is achievable through technology, we will shift the standards by which educational effectiveness is measured, which will create pressure on institutional providers to adopt mastery-based approaches even within the traditional system." That level of specificity suggests a genuine understanding of ecosystem dynamics rather than a general aspiration to scale.
BeMoreeDriven's Systems Change Thesis
At BeMoreeDriven Capital, we have developed a systems change investment thesis that runs alongside our sector-specific investment theses in enterprise AI, B2B SaaS, fintech infrastructure, and health technology. We are looking for companies that meet our conventional underwriting criteria — NRR above 120%, gross margins above 70%, demonstrated CAC efficiency — while also exhibiting the ecosystem impact signals that characterize genuine systems change companies.
This combination is rarer than either property alone. Many high-NRR SaaS businesses are excellent investments but have no systems change dimension — they are capturing value within existing ecosystems rather than restructuring them. Many mission-driven companies have genuine systems change ambition but have not yet demonstrated the commercial fundamentals that allow us to underwrite them with conviction. The companies that combine both — that are driving ecosystem transformation while building financially sound businesses — are the ones where we want to concentrate our capital.
The sectors where we find this combination most frequently are those where the system being changed involves significant access barriers — where the incumbent structure of the market has historically excluded large populations of potential participants because of cost, geography, or credentialing gatekeeping. Education, professional services, financial services, and healthcare are all sectors where AI is creating the conditions for this type of systems change at a scale and speed that was not achievable in previous technological eras. The investment opportunity in systems change companies in these sectors, at this moment, is one of the central convictions that shapes our current deployment strategy.
The world does not need more incremental products. It needs — and the market is beginning to reward — companies that understand the ecosystems they operate in deeply enough to change the rules rather than simply play the game better. That is the investment we are looking for. And when we find it, we intend to be the most patient, committed partner the founder has ever had.