Why Your Method Needs to Change as You Level Up

Most learners fail not because of bad advice but because they cling to the wrong method for their stage. Here’s how to match your approach to your level.

Created by the author using ChatGPT

My Story

I started taking French seriously about a year ago, after a few years of half-hearted effort before that. And the thing I learned isn’t really a single trick. It’s a model. I figured out that language learning happens in three stages, and each stage needs something different from the one before it.

Here’s the part that took me an embarrassingly long time to figure out: the real problem isn’t bad advice. It’s mismatched timing. Sentence-mining your way through Anki is brilliant at A1 and far less useful at B1. A grammar workbook earns its keep at the beginning but has diminishing value (unless you are targeting specific pain points) once you have more language under your belt.

People don’t fail at language learning because they picked a bad method. They fail because they keep using that method long after it is effective for their current stage. When this happens, progress stalls, learners get discouraged and the majority of them give up. It doesn’t have to be that way!

If you’re brand new, the first stage is where you are right now, and it’s worth knowing the shape of what’s coming. Knowing what you have to look forward to can help you push forward to get to the fun stuff! If you’ve been stuck at the same level for a while, there’s a decent chance you’ve been running the last stage’s playbook in the wrong place.

So instead of another list of “best techniques,” here’s a three-stage model, and what each stage actually needs from you. This isn’t about what feels productive, and it isn’t about what worked for someone else. Stage one is the one nobody likes talking about, so let’s start there.

A1–A2: Foundations

There’s no nice way to say this. The first stage is a grind, and honestly, there isn’t an elegant way to get around it. It is doing some important, foundational work, but it is also the part of language that feels most like you have re-entered high school. It isn’t fun or engaging, and you will spend a lot of the time you are at this level wishing you could be anywhere else. Here is the hard truth: if you skimp on the foundations at this stage, you are going to have to come back and fix things when you are at the intermediate level, so it is worth doing this stage correctly.

At this stage, you are building both a sound system and a basic grammar skeleton from nothing, and there isn’t a shortcut that skips the discomfort of not understanding much of anything yet. Trust me, I tried to find one when I started learning German!

A few things can make this stage slightly less miserable, though.

Photo by Jerin John on Unsplash

First: audio. Not just text — actual sound, from day one. If you only ever see your target language written down, you’ll learn to read a language you can’t recognize when someone speaks it to you, and if you try to speak to others, they won’t know what you are saying. You will be mapping your native language’s sounds to the target language’s words, and that will go about as well as it sounds like it will. Pair every new structure with how it actually sounds.

Some useful tools for this stage are the IPA pages on Wikipedia. I explain more about this process in my article The ‘Invisible Wall’ Blocking Your French Listening Skills. Yes, there is French in the title, but the information in this article is actually pretty language agnostic, so if you want a deep dive on the theory, this is the place to find it. This should be every learner’s first step when learning a new language.

Photo by Nick Morrison on Unsplash

Second, once you have reviewed the sounds, find yourself an entry-level course. There are many free resources online, so do a search for “Free language courses + your target language” and that should yield some good results. Pick an actual course that has some structure to it so that you can build your foundation. Check out a few, maybe read the FAQ on the subreddit for your target language if there is one, and find one you like with audio and video as part of the course.

The next and arguably most important step is to commit to it. Do something every day, even if it is only 15 minutes. Do NOT resource hop, trying to find something better, more interesting, or that sucks less, because the fact is this stage sucks no matter what your resource. Most courses can get you through this section, but you have to commit to a course and stick with it to see actual progress. You can supplement with a couple of other things that I am going to mention below, which are (slightly) more fun, but going through the course is essential for a good foundation.

Photo by Daniel Polo on Unsplash

Third: This is also a good window for some preliminary “shadowing” to get your mouth in on the action. Find short, simple, A1 “story” videos aimed at beginners on YouTube and just copy what you hear, sound for sound, even before you understand all of it. You’re training your mouth and your ear at the same time, and you are doing it before either one has much to work with. French is especially unforgiving if you skip this — liaison, nasal vowels, and the way half the letters on the page stay silent mean that text-only learners can read a sentence perfectly and not recognize the same sentence the moment someone speaks it out loud. Get the sound in early, or you’ll end up having to cycle back and relearn it later, and it will be significantly more difficult then.

Created by the author with ChatGPT.

Fourth: Anki belongs here, too, but not in the way most people set it up. Skip single-word cards and use whole sentences. Make sure your cards have audio, and pictures are also helpful. If you find a 1000-word deck, make sure it has all of these components. There are free ones for some languages on Ankiweb. A word sitting by itself on a flashcard tells you almost nothing about how it actually behaves — what it pairs with, where it sits in a sentence, what it sounds like next to the words around it. A sentence gives you all of that for free, every time you flip the card.

Photo by Perfecto Capucine on Unsplash

Fifth: Start reading something every day. This is the place for graded readers or graded content. I really like Wirelingo and recently discovered Lenguia, which I also really like, but your library might have some graded reader options for free. Don’t get any paid resources too early, though, because they function based on once you are hitting the A2 stage. Depending on your target language, you also might be able to grab any content online and paste it into an AI with these instructions:

Take the news article below and turn it into graded reading material for a learner of [TARGET LANGUAGE] at level [LEVEL]. Please create a simplified version of the article keep the main facts and ideas. Use vocabulary and grammar suitable for [LEVEL]. Keep it natural, not childish.

Note that not every language will yield good results with this method, but most of the larger ones will.

None of this is supposed to feel good yet. That’s the part people forget, and it’s why so many learners quit right here — they expect Stage One to feel like progress, and it mostly feels like effort with very little to show for it. It feels that way because that is what it is. Building foundations isn’t sexy or fun. It’s just necessary to build anything else. You’re building something that you can’t see yet.

By the time you’re genuinely at A2 (reading, hearing, and writing the basics and beginning to do it without translating everything in your head first), that foundation work is mostly done, and you start to feel successful. What comes next requires a completely different approach.

A2–B1: Starting your output engine

This is where the model most people were using breaks quietly. By A2, you have added enough vocabulary and grammar that you feel like you are finally making progress. Then somewhere between A2 and B1, that progress flattens out, and the reason is almost always the same: only two of the four skills are getting any attention.

Once you’ve cleared A2, all four paths (speaking, listening, reading, writing) need to be in rotation, and not as an afterthought to whatever input you’re already doing. Specifically, output needs to take up about half your time. Not a fifth. Not “whenever I feel ready.” Half, split between speaking and writing. I generally recommend starting with talking to yourself using the structures you have been learning and writing every day, even if you start out with only a few simple sentences. Spoiler alert: fill-in-the-blank exercises do not count as writing.

This is the part everybody resists, including me. Output is particularly uncomfortable at this level in a way that input simply isn’t. It is uncomfortable for precisely the reason it is essential. When you’re reading or listening, a mistake is invisible. If you misunderstand something, you can usually get enough context from the structures surrounding it to continue on and feel productive. Nobody around you can see your gaps, and neither can you. When you’re speaking or writing, every gap in your language is right there, audible or visible, with your name on it. This is why people keep doing more input, because it feels like progress and doesn’t carry the same risk of embarrassment, and then later they wonder why their comprehension is improving, but they still can’t hold a conversation.

What actually convinced me to push through that discomfort was realizing that my lack of progress in those areas was a direct result of a lack of time spent on those areas.

This is going to sound simplistic, but it is also true. If you want to speak better, you must speak. If you want to write better, you must write. It is both more difficult and simpler than I expected. Fortunately, writing supports speaking more directly than I thought it would and grants you the luxury of thinking time when you are starting out. When you’re mid-conversation, you don’t have time to think, so you grab whatever structure is closest to automatic, even if it’s wrong, because the conversation isn’t going to pause for you to get it right. Writing doesn’t have that time pressure. You can sit with a sentence, notice that you’re not sure about a structure, look it up, and use it correctly. Do that enough times, and the correct version starts to be the one that’s closest to automatic, instead of the wrong one you’ve been defaulting to in real time.

This is a scaffold not a crutch. Use as necessary to help you get started but get started! Infographic created by the author with ChatGPT.

This is going to be uncomfortable for the same reason that it is effective. You will notice your gaps, and you will work to correct them. That discomfort means you’ve found the actual work, not that you’re doing something wrong.

One clarification, because people tend to swing too hard once they hear “50% output”: listening and reading don’t disappear at this stage, they just stop being the whole plan. Input is still what’s feeding your speaking and writing — it’s where the vocabulary and structures you’re about to produce actually come from. The shift isn’t input versus output. It’s making sure output finally gets its share of the effort, after spending Stage One getting almost none of it.

B1–B2+: Learning IN the Language

The real gear-shift happens at the edge between B1 and B2, and it’s the one almost nobody tells you about. Up to this point, the task has been learning the language itself. From upper B1 onward, the task is learning something else in the language.

Find a subject that actually interests you, and go consume it in your target language. Not “content about general topics” — a specific thing you’d want to know about, even if it were in your native language. History, cooking, true crime, urban planning, whatever it is. If there’s a MOOC available in your target language on that topic, even better; you get sustained exposure to one register and one set of vocabulary instead of skipping across a dozen unrelated topics a paragraph at a time.

Created by the author with ChatGPT.

This is also where you quietly graduate out of material made for learners. Graded readers and beginner podcasts did their job, but they were always a simplified version of the language, built around someone else’s idea of what you should know by now. Content made for actual speakers, about a subject you’d seek out anyway, doesn’t simplify anything for you. That’s exactly the point. You stop being managed by the material and start using it.

This is where workbooks stop being useful, and what I affectionately call JIT grammar should take over. JIT, as in just-in-time. You stop working through grammar topics in the order that some textbook decided you needed them. Instead, while you’re reading or watching something you actually care about, you’ll keep running into a structure you can read and basically understand, but couldn’t produce yourself if you tried. That’s your cue. You go look that specific structure up, right then, because you’ve already met it doing real work and you actually need it — not because it’s “grammar week seven.” Learning a structure because you’re stuck on it, in the middle of something you care about, sticks in a completely different way than learning it because a syllabus said today was the day. The trick is staying curious about those gaps instead of feeling bad about them.

Output still has to be 50% of your time here, too, ideally split fairly evenly between speaking and writing. The mistake people make at B1 is assuming the gear shift toward content means they can ease off output, since they’re “doing so much input” through their subject of choice already. It doesn’t work that way. Comprehension keeps climbing without your help at this point. Production doesn’t.

One more thing worth saying out loud, since it changes how I plan my reading time: not all input pulls equal weight for vocabulary. If I had to rank it, fiction comes first, then nonfiction, then video and television. Novels throw a wider range of vocabulary at you per page than almost anything else, and because that vocabulary is attached to a character doing something you care about, it tends to stick rather than slide past you. Nonfiction is good too, but it tends to circle a narrower, more domain-specific set of words once you’re past the first chapter. Video and TV are the most enjoyable of the three and worth doing regardless, but the vocabulary tends to skew higher-frequency and more repetitive — useful for fluency, less useful for actually expanding your range.

So, Where Are You, Actually?

Photo by Alina Grubnyak on Unsplash

Each of these stages asks for something almost opposite from the one before it: white-knuckle endurance, then forced output, then curiosity-driven grammar chasing, layered under more output. None of them work if you’re running the wrong one for where you actually are.

So it’s time for an honest gut-check. If most of your study time is reading and watching things you halfway understand, and you still can’t hold a basic conversation, you’re probably still in Stage One even if you have been studying for a while. If you can’t read content made for native speakers and understand about 80% of what you read, you are likely still at stage one. Get some graded readers (yes, readers, not a television series) and start reading more. Remember that fiction is your richest mining zone for vocabulary. Series books are particularly effective.

If you freeze the second someone expects you to actually produce your target language, instead of just following it? That’s Stage Two, whether you want to answer or not. And if conversation feels comfortable but something’s stalled anyway, you’ve probably outgrown your workbook without noticing — pick a subject, start chasing the grammar that keeps tripping you up, and let Stage Three do its work.

None of this is about being further along. It’s about matching what you do to where you are, instead of borrowing a method that worked for someone at a completely different stage — including, sometimes, an earlier version of yourself.