The first lesson with a new student

The first lesson is a sales meeting that neither party admits is a sales meeting. The student is deciding whether to spend several hundred pounds over the next six months; you are deciding whether you can help them and whether you want to.

Most first lessons are spent on the student’s history with the language — where they studied, what they have forgotten, which apps they abandoned. It is a comfortable conversation and it predicts almost nothing.

Thirty minutes, structured

0:00   Why now, and what for                     6 min
0:06   Get them speaking                          8 min
0:14   Diagnostic — level and pronunciation       6 min
0:20   Teach one thing, properly                  6 min
0:26   Plan, expectations, one task               4 min

The proportions matter more than the exact numbers. Note that speaking begins at minute six and the teacher’s own credentials never come up.

1. Why now, and what for

Two questions, not the usual five.

“What do you want to be able to do that you can’t do now?”

Push for situations, not levels. “Reach B2” is a proxy the student has adopted because it sounds like an answer. What you need is:

• Handle a work call without writing a script first
• Talk to my partner’s parents without them switching to English
• Read a novel without stopping
• Not feel stupid in the team lunch

Each of those implies a completely different syllabus. The last one is not a language goal at all, which is useful to know early.

“Why now?”

There is nearly always a trigger — a job, a move, a relationship, a deadline, a specific humiliation. The trigger tells you how much time they will realistically put in and how long the motivation will last. Someone relocating in four months behaves entirely differently from someone who has always meant to learn Italian.

Six minutes. Take notes visibly; it signals that the answers will be used.

2. Get them speaking, early

Within the first ten minutes, before you have diagnosed anything.

Not a test — a real question they can answer at their level, about themselves. “Tell me about your week” works for almost anyone above A1.

Three things happen at once. You hear their actual production rather than their self-assessment, which is usually wrong in both directions. They discover that this lesson involves them talking, which is what they came for and often not what they expected. And they get past the first-minute terror, which otherwise contaminates everything after it.

Do not correct anything yet. Let it run. Take notes.

3. Diagnose

You are looking for three things, and you can get all of them from six minutes of speech plus one short reading.

Level, honestly. Self-reported level is unreliable — people who studied for four years at school say “intermediate” and produce present tense only, and people who lived in the country for a year say “beginner” and are fine. Listen to what they actually build.

Pronunciation, specifically. This is where most first lessons fail. Do not just note “good accent” or “strong accent”. Identify two or three specific problems, and check whether they are perception or production.

The ninety-second version: pick a contrast you expect to be hard from their first language — our language guides list them for 29 languages — and run six minimal pairs as a listening test. Near chance means perception, which is a different intervention entirely. This single test will tell you more than the rest of the lesson.

What they avoid. Listen for the constructions that never appear. A student who never uses the past tense in six minutes of talking about their week is not using it because they cannot, and they have arranged their entire speech to route around it. The gaps are more diagnostic than the errors.

4. Teach one thing, properly

Six minutes, one item, done thoroughly enough that they leave able to do something they could not do at the start.

Pick it from what you just heard — a structure they reached for and failed, or a sound that will cause them trouble for years. Not from a syllabus, and not the thing you happen to enjoy teaching.

This is the part that sells the lessons, and it does so by demonstration rather than description. A student who leaves the trial able to produce a sentence they could not produce at the start has evidence. A student who leaves with a description of your methodology has a claim.

Then let them use it three or four times, in their own sentences, about their own life.

5. Plan, expectations, one task

The plan. Three or four bullets of what the next two months look like, connected explicitly to the goal from minute one. “You want to handle work calls. So: the twenty structures that carry a call, phone-specific listening, and the /v/–/w/ contrast, because that one is costing you on the phone specifically.”

Expectations, stated plainly. How much work between lessons, what you will correct and what you will let go, how long this actually takes. Under-promising here is the single best retention move available, because the students who quit at month three are overwhelmingly the ones who expected month-one progress to continue linearly.

Say the honest thing: progress in language learning is not linear, month three feels like standing still, and that is the point at which most people stop.

One task. Small, bounded, obviously finishable. “Record yourself saying these five sentences and send them before Thursday.” Two minutes of work. See homework that gets done for why the bounded ones get done and the open ones do not.

Notes, sent immediately. Whatever you covered, in a form they can reuse — sentences with translations, the two corrections, the one sound with a transcription. This is also a demonstration: it shows them what every lesson will produce, which is a considerably stronger argument than saying you take good notes. Format in lesson notes students reuse.

What to skip

Your qualifications. They already booked. Nobody has ever been retained by a CELTA.

Long grammar diagnostics. A gap-fill test tells you what they can do with time and no pressure, which is not the skill in question.

A full needs analysis form. Six questions in conversation beats a twenty-item questionnaire, and the questionnaire signals bureaucracy in the first five minutes.

Talking more than they do. Rough target for a first lesson: the student speaks 60% of the time. Most teachers overshoot on their own talking precisely because they are nervous about being judged.

Free or paid

Both are defensible. Free lowers the barrier and increases no-shows; paid filters for seriousness and loses some good students who were merely browsing.

A reduced-price trial captures most of the benefit of each and is what most established teachers converge on. Whatever you choose, thirty minutes is usually right — sixty turns into a free lesson, and the second half rarely changes anyone’s decision.

The thing that decides it

Students come back when they can tell they progressed, and they judge that on evidence rather than feeling.

Which is why the notes matter disproportionately in a first lesson. A student who leaves with five sentences they could not say an hour ago — written down, translated, with the pronunciation marked and audio attached — has something concrete. It is also the format they will get every week, so it doubles as the product demonstration.

The mechanical part of producing that during a lesson rather than after it is what Teachee handles: the sentences you capture arrive already transcribed, translated and phonetically annotated, so the student walks away with the artefact at the moment the call ends rather than the following evening.