From B2 to running therapy sessions in French

Pedro P. is a psychoanalyst. On the CEFR scale he was between B2 and C1 in French. He could already use the language. What he still needed was listening comprehension and oral fluency: the French people actually speak, and the French he could produce at that speed.
For about six months he used a voice tutor very intensively, about four hours a day. The stored French conversations add up to about 500 hours. In June 2026 he reached his goal and graduated from the app.
He put the reason in one sentence. "He alcanzado el objetivo de perfeccionar mi nivel de francés, así que ya no necesito la herramienta, que me fue extremadamente útil para ese propósito." In English: "I have achieved the goal of perfecting my level of French, so I no longer need the tool, which was extremely useful to me for that purpose."
The goal
The target was not a beginner course and not a certificate. It was enough spoken French to be competent in rooms where French is the working language: a session with a patient, a book club, a seminar. Listening and oral fluency were the two skills he named. Reading a text was not the shortage.
The method
The hours were habits, not a reserved study block.
- About an hour and a half of walking, talking about whatever was actually on his mind.
- Reading a French text aloud, then asking linguistic questions about it.
- Translating into French, out loud, from Spanish, English, Portuguese, Italian, or Catalan.
- Turning television subtitles into spoken French.
- Describing what he was doing while he cooked or put a room in order.
- Thinking out loud about a clinical case or a philosophical point.
Added up, that came to about four hours a day. The State of Speaking Practice report, for 20 August through 18 November 2026, puts the median ISSEN conversation at 4.3 minutes and the average at 6.3. His day was many of those stretches, attached to a walk or a chore. The same report finds the busiest hour, in the learner's own timezone, at 9 pm.
What he valued in the tutor, in his account: fast linguistic answers, "not as exhaustive as dictionaries and encyclopedias, but much faster and more practical"; an interlocutor for thinking out loud; a sense of humor that could follow a word game or a bit of role play; and "the simulation of an intelligent, cultured interlocutor," including that "it even accepts that the user knows more about something."
The result
Four things he does in French now:
- He sees psychoanalysis patients in French. He also sees patients in English, Portuguese, and Spanish.
- He is in two book clubs with native French speakers.
- He is in a French conversation workshop.
- He takes part in frequent seminars, conferences, and working groups in French.
"The tool helped me reach the level needed to be competent in all those spaces." And, if he ever wanted the same kind of help in another language: "If in the future I needed a tool for another language, for example German, I would certainly consider ISSEN first."
What changed since his feedback
In January he asked for faster speech, clearer transcription, less repetitive openings, and fewer turns that ended on a question. Speech speed now goes up to 1.5x, which is higher than the January maximum. Transcription now runs on newer, more accurate speech models. A reply does not have to end with a question, and a later turn does not have to open the same way the session does. Listening to French at a native pace is still the main gap: translate subtitles out loud, and keep fast native speech beside the tutor, in the workshop, the book clubs, and the conferences.
What an advanced learner can copy
Attach the practice to a cue you already have. James Clear calls this habit stacking: after an existing action, do the new one (Clear, 2018). B. J. Fogg's version is the same shape, an anchor in a routine you already complete (Fogg, 2019). The walk and the kitchen were the anchors. A reminder to "study French" was not.
Make yourself produce the sentence. Merrill Swain's output hypothesis says that speaking and writing do work that input alone does not. You notice the hole when you try to say the thing, and you test whether the form you chose actually works (Swain, 1985; Swain, 1995). Translating a subtitle, or explaining a case out loud, is that test.
Ask for linguistic correction, and stop there. At advanced level, tutor-initiated corrections are limited to major mistakes, one at a time, and they are supposed to be sparse. There is also a student-initiated setting: the tutor waits until you ask. A narrow prompt fits this kind of practice. "Correct only the French. Don't comment on the idea." Then one real question: why this preposition, why this tense, is this register off.
Sources
Clear, J. (2018). Atomic Habits. Avery.
Fogg, B. J. (2019). Tiny Habits. Houghton Mifflin Harcourt.
Swain, M. (1985). Communicative competence: Some roles of comprehensible input and comprehensible output in its development. In S. M. Gass and C. G. Madden (Eds.), Input in second language acquisition (pp. 235-253). Newbury House.
Swain, M. (1995). Three functions of output in second language learning. In G. Cook and B. Seidlhofer (Eds.), Principle and practice in applied linguistics (pp. 125-144). Oxford University Press.