Spend the first month on sounds and on a few hundred words, and do not study grammar. Everything is new in the first month, so everything you do produces a visible gain, and the correct use of that month is to build the raw material the second month will need.
The second month is where almost everybody stops, and it is worth knowing the shape of it in advance, because knowing that it is coming is most of what gets people through it.
Eight decisions for the first three months
- pronunciation first, before habits set. Sounds learned wrong in the first month take years to correct, and the first month is the cheapest time to fix them.
- the first thousand words are function words and everyday nouns, and they carry an extraordinary share of ordinary speech.
- learn phrases before grammar. A stock of whole sentences produces speech immediately; a stock of rules does not.
- one course, finished. Starting four and abandoning each at chapter three is the commonest failure at this stage.
- fifteen minutes daily beats two hours on a Sunday, because the memory works on frequency rather than on volume.
- speak from week one, badly, out loud, to yourself if necessary.
- expect three months to A1 at that pace, which is a real and modest target.
- do not skip the alphabet sounds. English spelling is unreliable, so learning the sounds separately is not optional here.
What the first three months actually contain
Sounds first, because English spelling will lie to you from the first word and a learner who has never heard the sound system will spend a year saying words as they are written. A fortnight on the vowels and on where the stress falls is worth more at this stage than any amount of vocabulary.
Then the highest-frequency words, which are a much shorter list than a beginner expects and cover a much larger share of ordinary speech than seems reasonable. And alongside them, from the first week, the phrases for managing a conversation: how to say you did not understand, how to ask for a repeat, how to buy three seconds.
Grammar comes third and it comes late deliberately. A beginner with two hundred words and no grammar can be understood constantly; a beginner with the present perfect and forty words cannot say anything at all. The rules become useful once there is material for them to organise.
The shape of month two in English
Two things arrive at once and they arrive for everybody. The words get longer, and once they do, English spelling stops being a rough guide and becomes an obstacle: a learner who was reading three-letter words comfortably meets thought, through and though in the same week and discovers the writing system does not encode the sound.
And the irregular verbs turn up. There are around a hundred in ordinary use, they are the commonest verbs in the language, and there is no pattern to learn. In month one every new thing was a rule; in month two the most useful hundred items are a list to be memorised, and memorising a list feels nothing like learning.
Underneath both, the app runs out. A free app's exercises are built for the first month, where a sentence is four words long, and by the second month real sentences are longer than anything the exercises contain. Progress stops being visible, the daily streak keeps going, and nothing measurable happens for about six weeks.
Where it stops: reading a flat stretch as failure
The stall is the interpretation rather than the plateau. Nothing has gone wrong at week eight. The gains have moved from things you can see in a session to things that show up over a month, and the learner has no way to see a month.
What they do have is a daily signal that has gone quiet, so the conclusion is that they have no aptitude, or that the method is wrong, and the response is to change method. Changing method in month two resets the small amount of accumulated familiarity and starts the same curve again, which is why some people have four month-ones and no month-three.
The measure that survives this is not a session. Record two minutes of yourself talking at the end of month one and again at the end of month three. That comparison is visible when nothing else is.
The app, and exactly what it is for
Duolingo is the default and it is genuinely good for the first month. It supplies a daily habit at a time when the habit is the hard part, it drills the highest-frequency words in the order that pays, and it costs nothing. Use it and do not feel bad about using it.
It replaced the beginner's coursebook and its cassette, which needed a class and a timetable around it to work at all. What it stops being enough for is month two, and the reason is structural rather than a criticism: its exercises are short because short exercises are what a daily habit can sustain, and the thing you need in month two is longer input than an exercise can hold. Keep it as the habit and add something else as the material.
What it looks like when it goes wrong
The recognisable failure is a long streak and no speech. The daily record is intact, three hundred days are logged, and the learner cannot hold a two-minute conversation, because nothing in the format ever asked them to produce anything under time pressure.
The second is starting again. A learner who stopped at week eight comes back six months later and restarts from the first unit, because that is what the app offers, and repeats the month that was already finished.
The sound work in the first fortnight is being understood the first time you say it, and the conversation-managing phrases that belong in week one are the forty phrases that carry a beginner through a real exchange.
What comes after month three is a longer version of the same problem, where the flat stretch is measured in years rather than weeks and has a different cause underneath it: why progress stalls at B1, and what moves it.