Fixing Ed-Tech’s Investment Model

(Illustration by iStock/Mykyta Dolmatov)

At least since the Clinton era, the education technology sector has often been hailed as a catalyst that will fundamentally transform global education systems. As of 2023, the market had ballooned to over $320 billion in revenue. But in the context of annual global education expenditures surpassing $6 trillion, ed-tech's economic contribution still appears underwhelming by an order of magnitude.

The lion's share of overall education spending is tied up in legacy cost categories. In US K-12 education, over 95 percent of budgets cover compensation, operations, and services, leaving just about 2 percent for technology. This is significantly lower than the 8.2 percent average tech spending in US corporations and far from the 22 percent in digital industries. Even then, two-thirds of purchased educational software licenses in the US remain unused. Ed-tech in pedagogy is often a fringe novelty and not the trusted method.

This does not prevent the industry from existing in a permanent state of excitement and hype: For instance, ed-tech is currently at its peak popularity as a google keyword, beating climate tech by a large margin and equaling med-tech. Yet, the private health care market capitalization is estimated at $5T, representing 60 percent of the $8T market size; the global education market capitalization trails far behind at about 5 percent of the $6T aggregate spend.

After a pandemic-driven wave of excitement, private ed-tech investment dropped by 50 percent to $25B in 2022. While ed-tech deals outnumber climate tech two-to-one, they attract on average only an eighth of the funding. This underscores a basic issue: Ed-tech's funding markets appear ill-suited to support effective learning solutions.

Learning the Wrong Lessons

Ed-tech, encompassing both school and workforce learning, has a unique challenge: the end user is far from the buyer. Ed-tech is almost always purchased by someone other than the learner, and frequently by someone other than the teacher. This disconnect can delay recognition of quality products and entrench subpar ones in long software review cycles. The "product-market fit" concept from other tech sectors, which assumes that addressing user problems will drive demand, can be deceiving—the best solutions will not necessarily dominate.

Add to this the inherent difficulty in measuring ed-tech efficacy, often requiring resource-intensive research projects that cost upwards of six figures and lacking agreed-upon, universal metrics. Even if ed-tech buyers do want to procure the best available products, knowing which solutions are best-in-class can be a herculean feat. This, in turn, makes it difficult for investors to predict adoption, and thus underwrite it.

Ed-tech is also characterized by its vast diversity. The market is segmented into areas like higher education, K-12, pre-K, corporate, and open learning, among others, each with their distinct sub-segments and unique requirements. This diversity makes it challenging for investors to draw direct comparisons: with over 30 sub-segments and a total addressable market (TAM) of $320B ($115B for the US), each ed-tech investment category averages less than $4B in size. For investment funds looking at 10-20 ventures, this size is limiting, preventing specialization.

The small total addressable market problem spells further trouble for the venture capital world. Apple informs that there are over 80,000 iPhone educational apps globally, while the number of US ed-tech companies is estimated to be more than 10,000. Combine this with the market size to conclude that an average ed-tech company’s expected revenue is $10M. This appears irreconcilable with the well-entrenched VC idea of expecting 10-fold investment returns and the notion of unicorn hunting. While the ed-tech industry has seen around 30 of these mythical creatures thus far, they remain a statistical oddity rather than a viable investment thesis.

From an investor’s standpoint, the characteristics of the ed-tech capital universe are not unlike those of the used car market studied by George Akerlof in his 1970 paper outlining “the lemons problem.” That is, there is a proliferation of investment targets, yet investors do not have a foolproof way of gauging their quality prior to purchasing equity stakes, nor can the founders communicate it reliably. Therefore, investors are positioned to end up with subpar, low-conviction investments: few are likely to be “peaches,” but most are likely to be “lemons.” This gradually drives the “peaches” out of the market; the ed-tech founders who have created products that could truly contribute to learning outcomes and hold potential for widespread adoption are priced out, as the expected “lemon” term sheets are unattractive. The equilibrium results in only “lemons” being funded, and potential “peaches” leaving the market with no funding at all.

The ed-tech “lemons problem" is evident in financial data. Ed-tech venture funds average a 17 percent return, while other tech VCs average 36 percent, following data from Preqin for all VCs that have invested in ed-tech at least once between 2008 and 2023. Many education-focused VC funds have closed, and few have exited, resulting in underperforming ed-tech assets in portfolios.

End users are increasingly noticing the issues with "lemons"—for instance, the UN recently released an alarming report highlighting that most ed-tech products are never evaluated, with less than 1 in 8 UK companies undergoing rigorous testing or disclosing third-party certifications. In the US, only 1 in 10 ed-tech decisions are backed by peer-reviewed evidence, and much of the data is from biased sources.

A recent survey by the State Educational Technology Directors Association found that more than half of US teachers don't think they effectively use available ed-tech, with fewer than one in 10 seeing a need for more. This "lemons" issue reflects a broader delay in the anticipated ed-tech revolution, as untested tools overshadow the promises of “peaches.”

Moving to an Impact-Centric Investment Model

While the influence of investors on today’s ed-tech issues is evident, they also have the potential to spearhead reform. Drawing on lessons learned from impact investing in other markets, there are four social innovations relevant to the ed-tech world that could address the underlying “lemons” problem:

1. Embed impact intentionality in the investment process.

Investors who aspire to deploy capital in ed-tech should seek out targets with the explicit assumption that, in the long haul, their capital will generate a higher total return if the technology creates real educational impact. To that end, they should aim to formally establish the efficacy of their investment candidate during the due diligence process. While feasibility studies or RCTs are prohibitively expensive in the minimum-viable-product stage, ed-tech efficacy can be reasonably understood using less resource-consuming techniques. The US Dept. of Education, for instance, offers a framework that classifies ed-tech solutions under four tiers. Level IV is granted for “promise of evidence” and “having a rationale,” while Level III requires “one correlation study with proper statistical controls.” Certification is granted through a government-sponsored institution, What Works Clearinghouse; investors can plug into such frameworks on the assumption that the best ed-tech products would not have significant trouble becoming certified, and if they do, perhaps they should not be scaled. An example of an ed-tech investor that leverages such logic well is Owl VC, whose annual impact report details the rigor of efficacy measurement for its portfolio companies, disclosing how many have a research basis for their product, conduct quasi-experimental studies, or have a full-time staff member dedicated to outcomes.

2. Understand the broader impact potential by reviewing academic literature.

Investors can supplement the bottom-up logic of evidence levels by integrating a top-down synthesis of broader academic research. The ERIC database alone holds over 28,000 studies on educational efficacy, a resource an investor can tap into for deeper comparative analyses of prospective ed-tech tools against existing interventions. By understanding the academic consensus—triangulating papers and making assumptions where data is unavailable—a comprehensive picture of expected efficacy can be crafted, addressing questions such as scalability, depth, or additionality of impact. Such diligence ensures capital is allocated to not merely functional solutions but to those that excel beyond current alternatives. Frameworks like Impact-Weighted Accounts can then help to translate the academic insight into quantifiable projections, offering benchmarks for future impact targets. An example of an investor who leverages such an approach in a different industry is Summa Equity, a private equity fund with a sustainability focus whose impact reports use academic research to quantify environmental impact of its investments and set targets for its portfolio companies moving forward.

3. Create an infrastructure for impact reporting on a portfolio level.

While a scaling start-up might shy away from building an efficacy measurement and reporting function (rightly so—they should not be checking their own homework), portfolio investors might be well-positioned to do just that. Not only do investors have an incentive to get a sober evaluation of their investments’ efficacy—for governance and steering purposes—but they also benefit from intellectual synergies for deploying such impact infrastructure across their portfolio, i.e. measuring across 10 companies instead of just one. To do that, investors can hire internal staff—indeed, many impact funds have begun hiring talent dedicated exclusively to impact efficacy—but they can also leverage external solutions, like LearnPlatform. With investors spearheading efficacy measurement, industry standards’ convergence will be accelerated, also creating a more refined understanding of educational impact in relation to commercial potential. In return, impact sophistication will enable an investor to better pick future investment candidates, allocate resources across existing portfolios more efficiently, and finally report impact to their own limited partners, expanding future access to capital.

4. Align hold periods with a realistic scaling timeline.

Finally, it is important to adjust the impatient logic that has plagued some ed-tech venture investments. The classic VC paradigm of spending money to accelerate sales does not apply to most ed-tech scenarios—sales cycles to tech-skeptical schools and expense-conscious corporates are significantly longer, and traction is built not virally but over multiple academic years. Investors who seek ed-tech targets should consider reframing their return expectations, moving away from the idea of “unicorn hunting” to the idea of “camel grooming”—holding businesses for seven years or more and helping them patiently build a client base while fine-tuning the product. Given the small market sizes and fragmented landscape, ed-tech businesses are unlikely to become unicorns either way. However, they can provide stable and recurring cash flow on sticky client bases, which offers a commercially attractive investment strategy if properly underwritten, especially in recession-jittery markets. By lengthening hold periods and adjusting the scaling ambitions, investors can create the necessary space for end-users to adopt and evaluate the products, decreasing the unhealthy pressure that the “lemon” sellers have introduced, inviting the “peaches” back onto the market.

The ed-tech revolution can still happen—after all, education is a $6T global spending category with a very meager software share. Just as buying a used car is easier nowadays than in the 1970s, the industry can redeem itself from the “lemons problem.” With the faulty funding market being one of the root causes of the present underdevelopment, a self-imposed reform initiated by investors can benefit all stakeholders—most importantly, the learners.

Read more stories by Jakub Labun.