The wrong scoreboard
The edtech industry has spent a decade optimising for numbers that feel meaningful but measure the wrong thing. Engagement — time on platform, clicks, sessions per week — captures attention, not understanding. Completion — the percentage of a course finished — captures persistence, not knowledge. We track these numbers because they are easy to instrument and easy to report. Not because they correlate with what actually matters.
What forgetting curves reveal
Hermann Ebbinghaus mapped the forgetting curve in 1885. We have known for over a century that the human memory system decays exponentially in the absence of retrieval practice. A learner who completes a module with a 90% quiz score and never revisits the material will retain roughly 20% of it ten days later. This is not a failure of motivation or effort — it is how biological memory works. Any learning system that ignores this fact is not a learning system. It is an attention system.
Any learning system that ignores the forgetting curve is not a learning system. It is an attention system.
Retention as a design constraint
At Arconite, we treat retention as a hard constraint, not an afterthought. Every product decision — from how Gripho structures a conversation to how Skemax sequences module content — is evaluated against one question: will the learner remember this in thirty days? This forces discipline. It rules out interaction patterns that feel engaging but generate no durable trace. It forces spacing, interleaving, and retrieval into the core experience rather than tacking them on as optional review features. The result is slower, quieter interfaces that feel less impressive in demos and work considerably better in practice.