You know the feeling. You downloaded Duolingo with the best intentions. You did your first lesson, earned your first XP, maybe even kept a streak going for a while. Then somewhere around week three or four, you opened it, tapped through a few exercises, and thought: why am I doing this?
You closed it. You didn't delete it — that would feel like admitting defeat — but you haven't opened it since. It's still there on your home screen, the small green owl quietly judging you.
If this sounds familiar, you are not alone, and you are not weak. The data on Duolingo retention is brutal.
This isn't an accident. It's a consequence of how Duolingo is built — and more importantly, what it optimises for. Understanding the gap between what it claims to teach and what it actually produces is the first step to learning a language in a way that actually works.
The engagement trap
Most language learning apps are, first and foremost, consumer apps. They are built to be opened daily, to drive session length, and to convert free users into paying subscribers. Fluency is the marketing promise. Engagement is the actual product.
These goals are not always in conflict. But they often are. And when they conflict, engagement wins.
This is why the dominant mechanic in Duolingo — and the apps that have copied it — is the streak. A streak is a brilliant retention tool. It creates a daily obligation — not to learn, but to open the app. It activates loss aversion (a powerful psychological force) in a way that has nothing to do with language acquisition. You're not motivated to learn Spanish; you're motivated not to lose your 47-day streak.
Researchers studying motivation in digital learning environments have observed a consistent pattern: extrinsic reward systems (points, streaks, leaderboards) reliably increase short-term engagement and just as reliably undermine long-term intrinsic motivation. The dopamine hit from a Duolingo streak notification is real. But over time, users begin to associate the app with obligation and mild anxiety rather than with the genuine pleasure of progress.
"When the reward becomes the goal, the original goal disappears."
— Deci & Ryan, Self-Determination Theory, 1985
This is the engagement trap: the mechanisms designed to keep you coming back are precisely the mechanisms that make you want to stop.
The content problem: why nothing you learn feels real
Beyond engagement mechanics, there is a deeper problem with most language apps: the content is fake.
Not fake in the sense of inaccurate — the vocabulary is correct, the grammar is valid. Fake in the sense that it bears no relationship to any real communicative situation you will ever encounter. "The elephant drinks the purple milk." "My uncle is a doctor who lives in a submarine." These are real example sentences from popular language learning applications.
Language acquisition research is extremely clear on why this matters. Stephen Krashen's Input Hypothesis — one of the most cited and debated frameworks in applied linguistics — holds that we acquire language not by studying rules, but by understanding messages in that language. The critical word is understanding: the input must be comprehensible. But Krashen and subsequent researchers have added an equally important qualifier: it must also be meaningful.
Meaningful, in this context, means that the learner actually cares about the content. That it connects to something real in their life. That it communicates something they want to know.
Translating "the elephant drinks purple milk" gives your brain no reason to remember those words. There is no meaning to anchor the vocabulary to. This is not a minor inefficiency — it is a fundamental failure of the learning model. You are not acquiring language; you are completing tasks that happen to involve words in another language.
The level mismatch problem
Even for apps that provide more realistic content, there is a second obstacle: the content is almost never at the right level.
Krashen called optimal input i+1: material that is slightly beyond your current level. Enough challenge to stretch you, not so much that comprehension collapses. It is a narrow band. Too easy and you're not acquiring anything new. Too hard and you're not understanding enough to acquire anything at all — you're just staring at words.
Most language apps respond to this challenge by creating fixed lesson tracks: you advance through levels in a predetermined order, regardless of your existing knowledge or how your level actually develops. You might be genuinely A2 in reading but A1 in speaking, or strong on present tense but weak on subjunctive — nuances the linear app curriculum cannot accommodate.
The alternative approach — providing real content calibrated to your level — is what reading-based language learning has always offered. The challenge, historically, was finding enough of it. Graded readers are limited in topic and scope. Self-selecting from the real internet meant either material that was too simple (children's content) or too hard (native-level news).
This is the gap that modern tools have finally started to close.
The boredom is the signal, not the problem
When you feel bored with Duolingo, most productivity advice will tell you to push through. Build the habit. Trust the process. The boredom is a sign you're not committed enough.
This is wrong. The boredom is a signal. It is your brain telling you that what you are doing is not producing real learning, and your instinct to stop is probably correct.
The research on this is actually encouraging. Boredom in learning contexts tends to arise from two sources: content that is too easy (under-stimulation) or content that feels irrelevant (lack of meaning). Both are fixable — not by adjusting your mindset, but by changing what you're reading.
Human beings are not naturally bad at learning languages. We are the only species that does it at all, and every neurologically typical person on earth has successfully acquired at least one language to native level with no formal instruction whatsoever. The boredom you feel when grinding through Duolingo lessons is not a character flaw. It is a reasonable response to an environment that is failing to provide what your brain needs to learn.
What the research says actually works
The most consistent finding in second language acquisition research over the past forty years is that extensive reading — reading large amounts of text at an appropriate level — produces robust vocabulary acquisition, grammar internalisation, and reading fluency in ways that isolated drilling and translation exercises do not.
Paul Nation's work at Victoria University of Wellington has quantified this with unusual precision. His research suggests that incidental vocabulary acquisition through reading — picking up words from context without explicit study — requires a reader to encounter a word roughly ten to twenty times in varied contexts before it becomes reliably known. This kind of repeated, contextualised exposure simply cannot happen through flashcard apps, which present words in isolation.
More importantly, extensive reading is self-sustaining in a way that app-based learning rarely is. Reading something you find genuinely interesting creates the conditions for what Mihaly Csikszentmihalyi called flow: the absorption in an activity that makes time pass without noticing. You are not completing a lesson. You are finding out what happened next.
This is the crucial difference. Motivation in Duolingo is extrinsic — streaks, points, badges. Motivation in reading is intrinsic — you want to know what the article says. Intrinsic motivation is more durable, more resistant to the Saturday morning laziness test, and more likely to survive into the second year of learning.
The practical problem with "just read more"
If reading is so effective, why don't more people just do it?
The honest answer is that finding the right reading material at the right level has always been genuinely difficult. The options have historically been:
Graded readers: Carefully levelled, but the content is either patronising or dull. "María goes to the market. She buys apples." Nobody who has read a real newspaper in their life finds this engaging for long. The vocabulary is controlled at the expense of everything that makes reading pleasurable.
Native-level content: Newspapers, magazines, websites in Spanish or French are exactly the kind of material that makes reading worthwhile — current, relevant, sophisticated. But for a learner below B2 or C1, most of it is incomprehensible. Rather than i+1, it is i+20. Rather than stretching your vocabulary, it just makes you feel inadequate.
Language learning podcasts and YouTube: Genuinely useful, especially at beginner levels. But audio-only input, without the reinforcement of reading the text, produces slower vocabulary acquisition. And most learners eventually want to read — the medium has its own value.
The gap — engaging, real-world content at your exact level — is precisely what has, until recently, been genuinely hard to fill. But the shape of that solution matters.
The personalisation breakthrough: reading exactly what you'd read anyway
Most "reading for language learners" tools approach the problem by curating a library: a selection of articles, pre-levelled, waiting for you to browse. It is better than drilling. But it still involves someone else deciding what is worth reading.
The more interesting approach — and the one that changes the learning experience fundamentally — is to eliminate that curation step entirely. Instead of picking from what's available, you read whatever you want. Whatever you would be reading in English anyway. The article about the football transfer you've been following. The piece on climate policy that caught your eye. The profile of the director whose new film you just watched. You paste the URL, and within seconds you have the same article in Spanish or French, at your exact level, ready to read.
This is a different proposition than "here are some interesting articles for learners." It is: your reading life, in your target language.
The implications for motivation are significant. The research on intrinsic motivation in language learning consistently finds that learner autonomy — the sense of genuine choice over what you engage with — is one of the strongest predictors of long-term persistence. When you choose the content, you are not completing someone else's curriculum. You are reading something you actually wanted to read, in a language you happen to be learning. The distinction sounds small. In practice, it determines whether you open the app tomorrow.
There is also a vocabulary benefit that goes beyond motivation. The words you encounter in an article about something you care about are words you will remember, because they are embedded in context that matters to you. The word for "transfer fee" or "climate pledge" or "cinematographer" sticks when you encountered it in an article about something real, that you read because you wanted to know what it said — not because it appeared on a flashcard.
And crucially: with Lectura, the same article exists simultaneously at A1, A2, and B1. You switch between them with a tap. If the B1 version is stretching you too hard today, drop to A2 and keep reading. If the A1 version feels too simple, step up mid-article. This kind of frictionless level-switching is new. Language learning has always required you to commit to a level and find the right material for it. Now the material adjusts to you.
The honest version of "what to do instead" is not a different app or a different platform with a better-curated library. It is a fundamentally different model: one where your language learning is not a separate activity you do for twenty minutes with a purpose-built tool, but the same reading you were going to do anyway — just in a language you're building.
Read what you actually care about — in Spanish
Or in French
What "reading at your level" actually looks like
The CEFR framework — Common European Framework of Reference for Languages — divides learner ability into six levels: A1, A2, B1, B2, C1 and C2. The first three are particularly important for most adult learners, because most people begin learning as adults and spend the majority of their active learning time between beginner and intermediate.
At A1 (beginner), you can understand familiar words and basic phrases. Vocabulary is around 500–1,000 words. Sentences are short and present tense dominates. The goal is comprehension — getting the meaning of the text, not every word.
At A2 (elementary), you can understand sentences and frequently used expressions related to areas of immediate relevance. Vocabulary is around 1,000–2,000 words. You can follow a clear news story with some guessing from context. This is where most learners find their first genuine pleasure in reading — you're reading real things, not invented exercises.
At B1 (intermediate), you can understand the main points of clear standard language on familiar matters. Vocabulary has grown to 2,000–3,000 words. You can read a newspaper article and understand most of it. This is the level at which language starts to feel like a genuine window onto another world rather than a puzzle to be solved.
The same principle applies whether you're learning Spanish or learning French — the levels mean the same thing, and the importance of reading content calibrated to your level is identical.
The sources matter too
One of the underappreciated aspects of reading-based learning is that the source of the content affects what kind of language you acquire. Reading technology coverage builds a different vocabulary than reading sports journalism. Reading political analysis produces different syntactic patterns than reading health reporting.
For learners who want to be able to engage with real-world Spanish or French media — watching the news, reading online — this means it matters which sources you're reading from. A diet built entirely on entertainment coverage will leave gaps in your comprehension of other registers.
Sources like the BBC and CNN provide the kind of general-register journalism that transfers well to broad comprehension. For learners targeting the UK media landscape, sources like The Telegraph or The Daily Mail offer more varied registers and vocabulary.
The key is variety — not just of topics, but of sources. Different publications write differently, and exposure to a range of registers is part of what separates a learner who can only understand one type of Spanish from one who can read anything.
Try reading instead of drilling
Convert any article from the web into A1, A2 or B1 Spanish or French. No flashcards. No XP. Just reading things you actually want to read.
Start reading free →A practical framework for moving beyond Duolingo
If you're ready to try a reading-based approach, here is a concrete starting framework based on what the research supports:
Week 1–2: Find your level honestly
Most people are one level lower than they think — and that's fine. Starting at A1 when you feel like you should be at A2 is not embarrassing. It means the content will feel fluid, you'll finish articles rather than abandoning them halfway, and your confidence will compound. You can always step up. It's much harder to step down once you've committed to content that's too hard.
Week 3 onwards: Read something you actually care about
This is the single most important variable. A language learner who reads sports journalism they find genuinely interesting will outperform someone grimly completing science articles they find dull, even if the vocabulary in the second category is "better." The brain learns from engagement, not from effort. Find the sport coverage, the culture writing, the technology journalism that makes you want to find out what happens next.
Month 2+: Use the Explore feed
One of the more underrated aspects of reading-based learning is discovering what other learners find worth reading. The Explore feed on Lectura aggregates articles being read by the community — a useful signal for what's actually interesting rather than just what's algorithmically prominent. It also introduces you to sources and topics you might not have selected yourself, which is good for your vocabulary range.
The consistency question
Every piece of research on language acquisition through reading emphasises volume and consistency over intensity. Twenty minutes of daily reading produces better outcomes than two hours once a week. The reason Duolingo uses streaks is that this is genuinely true — daily exposure matters. The solution is not to abandon daily habits, but to replace the hollow streak mechanic with something that makes you actually want to read every day: a topic you care about, at a level where you feel competent and stretched simultaneously.
The honest comparison
Duolingo is not worthless. For building core vocabulary at the very beginning of learning, for introducing grammar concepts, for maintaining very basic daily contact with a language, it has genuine uses. If you've never studied Spanish before, ten minutes with Duolingo will teach you more than ten minutes of staring at a B1 newspaper article.
The problem is not that Duolingo exists. The problem is that it has been marketed as a complete solution, when the research clearly shows it is, at best, a starting point. The most common language learning trajectory — Duolingo for a few months, then plateau, then quit — is a direct consequence of people staying on the app long past the point where it can help them.
The uncomfortable truth is that reaching genuine fluency in Spanish or French requires consuming large amounts of that language at your level. There is no shortcut. The question is only whether you do that consumption in a way that feels like work — drilling exercises in an app — or in a way that feels like reading: self-directed, genuinely pleasurable, and built entirely around what you actually want to read.
That last part is what has changed. Language learning personalisation used to mean choosing which pre-built course to follow, or which curated library to browse. It now means something closer to its literal definition: your content, your topics, your sources, your level — switching between them with a tap as your ability grows. The same article you would have read in English anyway, waiting for you in Spanish or French the moment you want it.
If you have been bored by Duolingo, it is because Duolingo was boring you. The language itself is not boring. Spanish has more speakers than any other language on earth. French is spoken on six continents. The journalism, culture, politics and sport available in both languages is vast, rich, and genuinely worth reading — not as a language exercise, but because it is interesting.
You just needed a way to read it at your level. Now you have one.