Before EdTech Can Prove Impact, It Must Prove Viability

The EdTech Viability Index helps us understand whether promising tools can be implemented, sustained and scaled

An EdTech tool may work beautifully in a study and fail quietly in a classroom. That is not necessarily because the evidence was wrong, but because impact depends on conditions: devices, connectivity, teacher capacity, cost, curriculum fit, or everyday school routines. Researching these conditions is what implementation science does. Implementation researchers complement impact studies by asking: what conditions must be in place for this tool to work in real classrooms, and to keep working once the initial study or pilot support is gone?

Viability matters for both pilots and scaling

In implementation science, the conditions for piloting a tool (implementing for the first time) and the conditions for scaling (implementing over time and across settings) are related, but they are not the same. This distinction is sometimes blurred, especially when some governments or schools move quickly with digitisation and jump from procurement right to scale-up. But before a tool is implemented in a classroom, basic questions about viability need to be asked.

Are there enough digital devices, and are they shared or individually assigned? If the software relies on personalised algorithms for individual use, how will it function when several pupils share a device? Does the school have reliable connectivity? If a tool only works online, what happens in classrooms without stable Wi-Fi? Surprisingly, some interventions happen without first checking these basic assumptions.

Several Viability Factors

Beyond these initial viability questions lie broader issues of scaling and sustainability. These include the full cost of the software and its implementation (not only the monthly subscription per pupil, but also the costs associated with devices, maintenance, connectivity, data management, and technical support). Costs will differ substantially depending on whether the model assumes children bring their own devices, schools provide devices, or governments fund infrastructure centrally.

Teacher training is equally important. In classrooms, an EdTech tool must work not only for learners but also for teachers, who need the confidence, time, and pedagogical capacity to integrate it meaningfully into their instruction. A tool that is effective in principle may fail in practice if teacher training and familiarisation phase are not provided.

This is why, as EdTech Hub colleagues wrote, that EdTech Implementation Research helps bridge product development and education research and navigate a contested EdTech landscape.

Why this matters

For implementation research to develop into a robust science, it needs a systematic approach and a shared set of metrics. In efficacy research, the methodological expectations are relatively well defined, partly because many of the dominant designs, such as randomised controlled trials, have been used for long time in other fields (eg medicine) and adapted to education.

Effectiveness research is less systematised because it moves from controlled conditions into real classrooms, where use is shaped by teachers’ decisions, school routines, infrastructure, learner needs, and institutional constraints. Classroom observations, interviews, co-design sessions and other participatory methodologies are typically used in effectiveness studies. These methods are also used in implementation research but for a different goal: effectiveness evaluations ask whether the tool produces the intended outcomes in authentic educational settings. Implementation evaluations ask whether the conditions are in place for the tool to be integrated into teaching and learning as intended.

For EdTech, this distinction is essential: efficacy and effectiveness look at outcomes — the impact — measured through experiments, classroom studies, or structured user evidence such as teachers’ reviews. But implementation science looks at the conditions that make those outcomes possible.

A tool may fail to improve learning because it is poorly designed, but it may also fail because there are not enough devices, connectivity is unreliable, teachers are insufficiently supported, the tool does not fit classroom routines, or the costs make sustained use unrealistic. To understand these aspects, evaluators need shared rubrics and indicators; otherwise, we are comparing apples and pears and failing to advance the field through systematic accumulation of evidence about what makes EdTech viable in real educational settings.

The EdTech Viability Index

The researchers at the ICEI (WiKIT) have been tackling this challenge for some time. In our newly published paper, we present a set of benchmarks operationalised as measurable and traceable indicators, so that implementation insights gathered by different research teams can be compared with greater rigour. Taken together, these benchmarks constitute the EdTech Viability Index.

The purpose of the Index is to ensure that viability is measured consistently and holistically. So, not reduced to a single factor, such as the cost of a software subscription or teachers’ technical skills, but understood more comprehensively. Are all the conditions in place for a tool to be implemented in ways that support the intended outcomes? Remember that these outcomes may be defined through standardised tests, such as foundational literacy or mathematics skills, or more innovatively, such as future-ready skills, learner agency, collaboration, or problem-solving.

In all cases, viability should be assessed in relation to the outcomes the tool is expected to achieve, for whom, and under what conditions.

With shared indicators, the field can compare different tools, contexts, and forms of use more meaningfully, while building a cumulative understanding of what counts as successful implementation.

Applying the Index into practice

The EdTech Viability Index is supported by a set of practical tools. At its core is a rubric that can be used at different points in the EdTech evidence journey. At the selection or procurement stage, it helps policymakers, investors, and education leaders judge which tools are most suitable for a given context. At the evaluation stage, it helps researchers assess whether the conditions for meaningful implementation are in place before a solution is tested for impact.

The paper outlines suggested methods for collecting data across each viability dimension:

• technical viability
• economic viability
• human viability
• system viability

Systematically collected data can support comparison studies of interventions and tools. And this can be used by various repositories or databases that rank EdTech solutions according to both their likely impact and their practical viability. The International Certification for Evidence of Impact in Education - EduEvidence - has already expressed interest in adding a viability score alongside the existing impact score in its global EdTech solution list.

The technical paper and scoring formula provide the methodological basis for this work, but the principle is simple: viability is not an optional extra.

Interventions rated poorly on critical dimensions should be considered insufficiently viable and should not proceed to further implementation or impact evaluation without remediation.

In other words, before we ask whether a tool works, we must ask whether the conditions exist for it to work.