The history of educational innovation is, in large part, a history of unfulfilled promises for the majority. Every new technology—educational radio in the 60s, television in the 70s, computers in the 80s and 90s, tablets in the Obama era—has arrived with revolutionary predictions. And each time, the real benefit has reached only a small fraction of students: generally those who already had advantages.

The historical pattern of educational innovation

5% is not an arbitrary figure. It is a conservative estimate of the percentage of students who have historically received real and sustained benefits from technological innovations in education. The rest are left out not for lack of merit, but for lack of access, infrastructure, teacher training, and resources.

5% Students who receive real benefit from EdTech innovation. 45% Rural schools in LATAM without reliable internet. 2.3x Greater technological gap in low-income areas.

What makes the 5% problem particularly pernicious is that it is not evident from the outside. EdTech startups report millions of users, coverage in dozens of countries, retention rates of premium users. But if you break down the data by socioeconomic level, geographic area, and type of institution, the pattern emerges: innovation benefits those who already have more.

Why current solutions don’t scale

The underlying problem is structural and has to do with the incentives that guide the development of EdTech products. Most companies in the sector are evaluated by investors according to market metrics: paying users, retention, lifetime value. These metrics naturally privilege the segments with higher purchasing power.

“Digital inequity is not a bug of the EdTech system; it is a feature of how we have designed it. As long as we measure success with market metrics, we will optimize for the market, not for impact.” — Daniela Reyes, Mentu Labs

The three systemic barriers

  1. Unequal connectivity: 45% of rural schools in Latin America do not have access to reliable internet, which immediately disqualifies most modern EdTech solutions.
  2. Gap in teacher training: technology without pedagogy is noise. Most implementations fail not because of technical problems, but because of a lack of support for teachers.
  3. Incompatible business models: high-quality premium solutions are inaccessible to the institutions that need help the most. The public sector can rarely compete in response time with the private sector.

The Mentu model: designed for the 95%

Teacher co-design workshop in rural Peru

At Mentu Labs, we design from the outside in. We start with the most challenging contexts—limited connectivity, scarce resources, high teacher turnover—and build solutions that work there. The logic is simple: if it works in Ascope, Peru, it will work in Mexico City. The reverse is rarely true.

Our commitment: 70% of our active projects work with public or non-profit institutions in high-vulnerability contexts. Not because it’s easy, but because it’s where the marginal impact is greatest.

A call for collective action

The EdTech ecosystem needs a different conversation. One that measures success not in valuations or premium users, but in learning for the 95%. That prioritizes depth over speed of deployment. That recognizes that true scale is not reaching millions of users, but changing the learning conditions of millions of students.

That requires changes in how philanthropists and governments invest, how evaluators measure impact, and how startups themselves conceive their mission. It is a collective effort. And at Mentu Labs, we believe it is the work worth doing.