A Big Number With a Quiet Caveat
On June 29 England's Department for Education put a figure on its homegrown classroom-technology industry: a 5.9 billion pound market, sized alongside 796.2 million pounds of investment and growth running at 8.6 percent a year. By the underlying sector data that turnover sits across roughly 1,071 companies employing more than 26,000 people. On its own that is a confident growth story for a country that has worked hard to brand itself an EdTech exporter. The number is the headline the sector wanted.
The caveat attached to it is the part worth reading. The same DfE message calls for clearer usage data, which is a polite way of saying the government does not actually know how much of this 5.9 billion pounds of technology is used, how often, or to what effect once it reaches a classroom. A market can be large and opaque at the same time, and England has just admitted that its is both. That tension is the real news here, not the topline.
Why the Evidence Gap Matters
For the executives who buy and sell this technology, the visibility gap is not academic. Schools spend hundreds of millions of pounds a year on tools whose impact on learning and teacher workload is asserted far more often than it is measured. When a national education ministry says out loud that it needs clearer usage data, it is conceding that procurement across the system has been running partly on faith. That is an uncomfortable thing to admit about a 5.9 billion pound market.
We think this is the maturing pain of an industry that grew faster than its evidence base. The first wave of EdTech sold on access and novelty. The second wave, now arriving with AI, is being asked to sell on proof. England's call for usage data is an early sign that buyers, prompted by government, are starting to demand the metrics that other enterprise software categories take for granted. Vendors that cannot show usage and outcomes will find the next procurement cycle harder.
The Machinery England Is Building
To its credit, the DfE is not only naming the problem. In partnership with the Chartered College of Teaching it has launched an EdTech Evidence Board pilot to explore how to build credible evidence that products work, and an Impact Testbed programme that lets schools and colleges trial tools and share real evidence of their effect on workload, outcomes and inclusivity. These are modest, slow-moving instruments, but they point in the right direction: toward a system that asks vendors to demonstrate value rather than assert it.
Dr Cat Scutt, deputy chief executive at the Chartered College of Teaching, captured the underlying logic when she argued that AI has huge potential in education but also huge risks, and that the only way to manage both is for the workforce to be confident and competent in safe, effective use. Evidence and capability are two sides of the same coin. A testbed that proves a tool works is wasted if teachers are not equipped to use it, and trained teachers are wasted on tools that do not.
A Template Other Governments Will Copy
England is effectively running a public experiment in how a government governs a large, fast-growing EdTech market it cannot fully see. Size it, admit the data gap, then stand up evidence boards and testbeds to close it. That sequence is replicable, and other education ministries watching the AI wave hit their own classrooms will likely borrow it. The alternative, buying at scale without measurement, is exactly the posture that produced years of expensive, unproven deployments.
For enterprise and education buyers anywhere, the lesson generalizes cleanly. Market size tells you how much money is moving, not whether it is well spent. The 5.9 billion pound figure is impressive and almost beside the point. The valuable disclosure is that a serious government looked at its own EdTech estate and said, in effect, we need to see this better before we trust it. Anyone signing technology contracts in 2026 should adopt the same humility.
The AI Wave Raises the Stakes
The timing of this candor is not incidental. England is pushing AI deeper into its classrooms at the same moment it admits it cannot see how existing technology is used. That combination should give policymakers pause, because layering generative AI onto an evidence base this thin compounds the visibility problem rather than solving it. If a government cannot measure the impact of a learning app it has funded for years, it has little hope of measuring the effect of an AI tutor it deploys this term without far better instrumentation.
That is the strongest argument for the evidence boards and testbeds to move faster than they comfortably can. The usual pace of public-sector evaluation, careful and slow, is mismatched to a technology iterating every few months. We would rather England slow its AI procurement to match its evidence than rush its evidence to justify its procurement. The 5.9 billion pound headline will keep growing. Whether the country can finally say what that money buys is the test that matters.
