America spends roughly $30 billion annually on K–12 education technology, yet student outcomes remain stubbornly poor. This is not due to a shortage of innovation; it is because the system rewards products that are well-marketed and familiar over those with evidence of improving learning. Bans on devices and growing skepticism toward ed-tech reflect legitimate frustration, but the real question is about evidence: what helps students learn, and under what conditions?
Researchers, educators, and entrepreneurs are developing tools grounded in learning science and tested in real classrooms. What is missing is a reliable mechanism to move proven tools to students at scale while keeping unproven ones out. Research and markets operate on separate tracks, each with its own incentives and timelines, leaving too many effective ideas stranded in labs and pilots while inferior products reach millions of students because they are easier to buy or backed by recognizable vendors.
A handful of examples show what the bridge between research and scale can look like when conditions align. Magpie Literacy, developed through the Reading Reimagined program at the Advanced Education Research and Development Fund (AERDF), translated research on the decoding threshold into adaptive digital tools and formed a distribution partnership with Great Minds to reach students at scale. CueThink, supported by AERDF’s EF+Math program to embed executive function strategies into math instruction, was acquired by Imagine Learning after rigorous classroom-centered development. A2i, a K–3 reading intervention, took nearly two decades of sustained federal investment and multiple grants before Scholastic acquired it and integrated it into a national product suite. These success stories are exceptions, each dependent on rare alignments of patient funding, educator collaboration, and commercial partnership.
The fundamental problem is the funding gap in the “messy middle” — the stage where evidence becomes a usable product. Philanthropic and government grants typically fund discovery and evaluation. Venture capital looks for revenue and growth. Neither is well-suited to fund the iterative work of turning promising research findings into a classroom-ready tool: building the product, testing it with educators, refining it through use, and generating evidence across diverse settings. Innovations with strong evidence lack capital to build a product; those with a product face pressure to show revenue before building evidence. Federal education R&D compounds the problem through misaligned incentives. The What Works Clearinghouse, created in 2002 to help schools identify effective programs, became better at evaluating research designs than at helping practitioners act on findings. Across federal contracting and grant-making, the system rewards methodological precision, procedural compliance, and continuity — not usefulness. Large contracts tend to favor incumbent organizations skilled at navigating procurement rules, shielding them from the competitive pressure that would otherwise drive innovation and improvement. Peer review panels further narrow what counts as legitimate research, excluding the interdisciplinary coalitions that effective applied R&D requires.
These structural problems ripple into school purchasing decisions. District leaders navigating procurement rules, funding restrictions, and risk aversion default to familiar vendors. The market signal becomes visibility and familiarity rather than evidence of impact.
Fixing this requires changes across funding, evidence infrastructure, and incentives. On funding, philanthropy, government, and mission-oriented investors must treat translation as a priority — staying engaged through implementation rather than exiting at proof of concept. Programs like the Institute of Education Science’s From Seedlings to Scale grants and the National Science Foundation’s Accelerating Research Translation initiative point in the right direction by treating translation as a capacity to be deliberately built, not something that happens automatically after research concludes.
Schools need more than “yes/no” verdicts on whether a program “works.” They need probabilistic guidance, identifying what is more likely to work for which students, in which settings, and under what conditions. And they need that information when adoption decisions are being made, not years later. Procurement rules and funding structures should require providers to continue measuring impact after their products reach classrooms, shifting the market reward from reaching schools to improving outcomes once there.
Researchers need recognition for translation work. Federal agencies need to measure success by whether funded research produces tools that reach students, not only by study quality. Intellectual property arrangements must be addressed from the start, creating structures that reward commercial partners for building effective tools while protecting the public value that public and philanthropic investment is meant to generate.
The $30 billion already flowing into education technology will not produce the learning gains students need as long as the market rewards visibility over impact. The bridge from research to classrooms has been built in exceptional circumstances, by people willing to do the hard work of translation. The challenge is to stop treating those successes as exceptional and to build the funding structures, evidence systems, and incentives that make them normal. The students who need these tools cannot afford to leave that to chance.
This post is an abridged version of an article originally published by AEI.
