A Smaller EdTech Market Isn’t a Better One

The excitement around educational technology startups has cooled—but it would be a mistake to assume that what remains is therefore better. Markets do not naturally select for what is educationally meaningful or ethically sound. They select for what survives. And survival, especially in a tightening funding environment, often reflects commercial fitness rather than pedagogical value.

The current contraction in EdTech should not be read as a simple “clearing out of the noise.”

It is something more ambiguous, and potentially more consequential: a reconfiguration of which kinds of products, practices, and assumptions become embedded in education at scale.

Source: Tracxn

When the market “cleans up,” what actually gets left behind?

It’s tempting to compare this moment to industries where poor-quality products are filtered out over time—where only the strongest, most useful solutions endure. But EdTech doesn’t behave like a typical consumer market.

In education, the end users (students) are rarely the buyers. Procurement is mediated through institutions, policy pressures, and increasingly, platform ecosystems. That means weak or even harmful products don’t necessarily fail quickly. They can persist—sometimes for years—if they align with administrative incentives, reduce costs, or simply integrate well with existing systems.

So the current downturn raises a sharper question: what exactly is being selected for?

• Tools that demonstrably improve learning?

• Or tools that scale cheaply across institutions?

• Systems that support educators’ professional judgment?

• Or systems that standardize and automate it away?

Without asking these questions directly, we risk mistaking consolidation for improvement.

From hype to infrastructure—without scrutiny

In the early 2020s, EdTech was characterised by experimentation. Many ideas were shallow, but the field was visibly in flux. Now, as funding tightens, the sector is shifting from experimentation to infrastructure. Venture capital has fallen to around $2.4 billion globally—down from pandemic-era highs above $16 billion—marking the lowest level of investment in roughly a decade.

A clear example of this shift is how investment is concentrating. Instead of spreading across thousands of experimental apps and platforms, funding is now directed toward fewer, larger bets—particularly AI-enabled systems and workforce-oriented tools that can integrate directly into institutional operations and demonstrate measurable efficiency or outcomes.

This is where the stakes increase. Infrastructure technologies—assessment platforms, learning management systems, AI tutors, workforce analytics—do not just support education; they shape its conditions:

• what counts as learning

• what gets measured

• how progress is defined

• which learners are advantaged or excluded

If low-quality ideas survive at this layer, they don’t just underperform—they redefine the system around their limitations.

The real risk: that “rubbish” becomes standard

The intuitive story is that downturns eliminate the weakest players. But in EdTech, there is a credible opposite risk: that the most adaptable to market pressures—not the most educationally sound—become dominant.

This can produce a particular kind of “rubbish” that is harder to detect:

• Technically sophisticated but pedagogically thin systems

• Data-rich but insight-poor analytics dashboards

• AI-driven personalization that optimizes engagement metrics rather than learning depth

• Efficiency gains that quietly displace human judgment

These products don’t look like failures. In fact, they look very well like progress. They come with dashboards, metrics, clean interfaces, subscribed customers, and long-term contracts. But their underlying assumptions about learning are often narrow and instrumental; they may be entirely unethical about how they use education data, and so on. Moreover, once embedded, they are difficult to dislodge.

Quality needs redefining—not just enforced

If this moment is to be meaningful, “quality” in EdTech cannot remain a vague proxy for usability or growth. It needs to be rethought more fundamentally.

A high-quality educational technology should be able to answer, clearly and convincingly:

• What conception of learning does this tool embody?

• What does it make easier—and what does it make harder—for educators and learners?

• What kinds of behaviours does it incentivise?

• What forms of knowledge does it privilege or ignore?

• How does it handle uncertainty, ambiguity, and difference?

• How does it handle data, privacy, ethics, cybersecurity, and designs with regard to children's needs and wellbeing?

This moves the conversation beyond performance metrics into epistemic and ethical design: how knowledge, judgment, and fairness are built into the system itself—and how robust and safe the system is.

Identifying what is actually beneficial

The opportunity in this downturn is not simply to back “better companies,” but to develop sharper criteria for what counts as beneficial in the first place.

Promising directions are already visible—but they require careful distinction from their weaker imitations.

• Systems that make learning processes visible, not just outcomes

• AI applications that are inspectable and contestable

• Platforms designed for inclusion and variability, not just average-case optimisation

This is exactly where audit work becomes critical. Initiatives such as the EDDS Institute are attempting to move the conversation beyond claims of effectiveness and into verifiable practice. Their approach treats EdTech systems not as neutral tools, but as socio-technical infrastructures that need to be examined, stress-tested, and crucially—held accountable.

An audit, in this sense, is a structured interrogation of how a system works in context:

• What data is being collected, and on whom?

• How are outputs generated, and what assumptions shape them?

• Where are the points of opacity or potential bias?

• How does the system alter teacher judgment, student behaviour, institutional decision-making?

• How safe are these systems?

• What design principles are used to adhere to children's human rights, needs and wellbeing?

Of course, audit work does not scale as quickly as product deployment, and it does not immediately translate into revenue. Which is precisely why it needs deliberate attention. In a contracting market, audit capacity becomes a form of infrastructure in its own right—one that determines whether what survives is genuinely valuable, or simply well-adapted.

If left to market forces alone, a smaller field does not mean a better one—making it essential to decide, through audit, what is worth keeping.

A narrower field, a sharper responsibility

The contraction of the EdTech sector does create an opening—but not the one the usual narrative suggests. It is not simply that the “bad” ideas disappear and the “good” ones remain.

It is that fewer ideas remain—and therefore, each surviving system carries more weight in defining the future of education.

That makes this a moment of selection with real consequences.

If we are not careful, the industry will not emerge purified, but consolidated around a set of assumptions that were never adequately questioned. The result will not be better education through technology, but more deeply embedded versions of its current limitations.

The real task now is not to celebrate that the field has shrunk—but to know and decide, with much greater precision, what has and what deserves to remain.