You didn't lose motivation, you lost the word
You learned "desafortunadamente" on Monday, used it twice inside an app, and by the following Monday it is gone, not faded, gone, as if you had never picked it up. The explanation most people reach for first is about themselves, not disciplined enough, too easily distracted, maybe just not built for languages. None of that holds up under any scrutiny. Forgetting a word a week after you learned it is not a personal failure. It is the default behavior of human memory, and it has been measured and charted for more than a hundred years.
The forgetting curve is not a metaphor
In the 1880s, the German psychologist Hermann Ebbinghaus ran himself through thousands of memorization trials using invented, meaningless syllables so no prior knowledge could help him, then tested how much he could still recall at fixed intervals afterward. The pattern he documented, now known as the forgetting curve, has been reproduced in one form or another across a century of later memory research, people tend to lose roughly half of what they have just learned within a matter of days or weeks, unless that memory gets reinforced somewhere in between. A single exposure to a word, however well it lands in the moment it is taught, is not built to survive a week of silence afterward. That is not a flaw in you, it is the baseline.
The practical version of this shows up constantly. Someone crams fifty new words the week before a trip, feels fluent on the flight over, and finds half of them missing by the second day, not because the trip disrupted anything, but because a week had simply passed. The forgetting curve does not care how motivated you were when you studied, it runs on elapsed time regardless of intention, and the only thing that bends it is contact with the word again before it drops.
One graded reader is not enough, even when the method is right
This is not unique to drilling apps, either. It happens with reading too, which is worth taking seriously because reading real, comprehensible stories is one of the more evidence-backed ways to build vocabulary in a second language. A frequently cited study by Rob Waring and Misa Takaki, published in 2003, tracked what happened to 25 target words after learners read a single graded reader. Three months later, the meaning of only one of those 25 words was still remembered. Words that had appeared fewer than eight times across the text were not retained at all. The ones that survived were not the ones explained most clearly, they were the ones the learner had bumped into repeatedly without necessarily noticing. Take a word like "madrugar", to get up early. Meeting it once in a glossary is not the same as meeting it across several chapters of a story where a character keeps doing it before dawn, which is closer to what eight real encounters looks like in practice.
That number, one surviving word out of twenty-five, is not an argument against reading, and it is not an argument for any particular app either. It is evidence against expecting a single exposure, whether that is one story, one flashcard session, or one grammar explanation, to do the full job of retention by itself. Vocabulary needs repeated contact, spread out over time, or it quietly reverts to zero no matter how good the material was the first time around.
What gap-fill drills get wrong, and what they actually get right
This is where the usual complaint about Duolingo-style apps needs to be more specific than it does not work. Repetition is not the problem. Repetition is, in fact, the only mechanism that has ever reliably pushed back against the forgetting curve. The problem is what gets repeated, and on whose schedule. Filling in a blank with "como" ten times in a row trains you to recognize a pattern inside that exercise, which is a different skill from retrieving the word on your own, inside a sentence you did not write, three weeks later with no multiple choice options in front of you. And because most gap-fill systems repeat material on a fixed schedule rather than one shaped by what you, specifically, are about to forget, they tend to over-review words you already hold solidly and under-review the ones that are days from disappearing.
Spaced repetition decides what gets reviewed, not only when
This is the part a well-built spaced repetition system is actually designed to solve. FSRS, short for Free Spaced Repetition Scheduler, does not review every word on the same three-day or seven-day cycle by default. It tracks how reliably you have recalled each individual word and recalculates, word by word, the point at which you are statistically about to forget it, so the review lands right before that drop-off instead of long after it has already happened. That precision is the difference between a flashcard deck that keeps vocabulary in place and one that mostly produces the feeling of having studied.
Context still has to do its share of the work
None of this means spaced repetition on its own would have rescued the 25 words in the Waring and Takaki study, because the words that survived there were not the ones reviewed most mechanically, they were the ones met repeatedly inside real sentences, attached to the stakes of actually understanding what was happening in the story. A word drilled only as an isolated flashcard, divorced from any sentence, and a word met by tapping it mid-story, seeing its meaning and pronunciation, and then running into it again two paragraphs later in a slightly different form, are not the same memory. The second one carries context with it, and context is part of what makes a word retrievable weeks later, not merely recognizable in the five minutes after you studied it.
What a routine built around this actually looks like
Put together, the research points to something fairly unglamorous, a word needs to be met enough times, closer to eight than to one, spaced out on a schedule that matches your own forgetting rate rather than a fixed calendar, and ideally inside sentences you are reading for their own sake rather than only inside isolated drills. No single feature does all three by itself, and no single week of effort does either. A daily ten-minute habit of reading something slightly above your comfort level, combined with reviewing the words that are actually at risk rather than everything you have ever logged, will outperform an intense weekend of cramming almost every time, because the forgetting curve runs on calendar days, not on hours of total effort.
This is roughly the gap HabloHack is built around. Stories are graded by CEFR level, from A1 through C2, so the unfamiliar word in front of you sits inside sentences you can actually follow instead of ones you are guessing at. Tapping any word shows its meaning, pronunciation, and an example without pulling you out of the page, and the vocabulary you collect, whether it comes from your own reading, an official list, or words other learners have flagged, goes into flashcards scheduled with FSRS, which tracks each word's own decay instead of reviewing everything on the same clock. None of that turns a single reading session into permanent memory. Nothing does. But it is built around the actual mechanism that makes a word stick, rather than around the feeling of having studied it.