What should a therapist actually do?

What is a therapist there to do?

To understand what you are living with in enough detail that it changes what you can do about it.

Most people arrive expecting an examination, and some part of them hopes for a name at the end of it, because a named thing feels more manageable than a vague one. What happens is slower. I ask what brought you. I listen to the order you put it in, where the account speeds up, which year gets skipped. That is not assessment for its own sake. The difficulty you came with has a shape you have had no chance to see from outside, and seeing it is what changes what you can do with it.

So expect attention of an unusual kind: slow, specific, interested in detail you would feel silly taking to a friend. Expect me to ask about a particular Tuesday rather than about the general pattern, because the feeling is still in the particular one. Do not expect me to know in that first hour what the difficulty is made of. I will have an early impression, and I will give it to you as an impression, which is not the same as a conclusion.

Will a therapist tell me what they actually think?

Yes, and asking for it directly is a reasonable thing to do.

There is a stubborn picture of the therapist who withholds, nods and returns your sentence to you with the verb changed. That is not the work. If the way you are handling something is costing you more than it protects you, I will say so, and I will show my working, so that what you argue with is the reasoning rather than me. Being treated so delicately that nothing is ever said is not kindness. It is an expensive hour of company.

The other half of that is restraint about what I cannot know. I do not know why your father went silent for a decade. I do not know whether your partner withdraws out of hurt or out of indifference, and neither, probably, do you yet. Where a motive is unestablished I keep it as one possibility among several and say that is what I am doing, because a guess dressed as a finding gives you a false certainty you then have to build on.

None of this is separate from the relationship it happens inside. Across studies of people treated for depression, a stronger working relationship with the therapist generally went with better results, although the influence runs in both directions: improvement also seems to raise how the relationship is rated, and the review’s authors are careful about what can be concluded from 10 studies of varying design.¹ The practical reading is narrow. How well the two of you work is part of the treatment, not the foyer in front of it.

Who leads the hour, me or them?

You bring the material; the therapist carries the shape of the hour.

Each half feels like it ought to belong to the other person, which is why the first few sessions can feel unsteady. You want to be led, told where to start, spared the work of deciding what matters most. You also want to keep control of what gets opened and when. Both wishes are sensible, and they sit in the same person pulling in opposite directions.

• Yours. What you bring, and what you would rather not bring yet.
• Mine. Where the hour goes, what we come back to, saying when we have drifted somewhere safer, and ending on time.
• Both. What the work is for, said out loud early and revised as it becomes clearer.

You do not need an agenda, a tidy history or a clear question. “I do not know where to start” is a start, and I will work from it. You also keep the right to say you would rather not go into something today. That is not resistance to be got around. It tells me how much is bearable in one sitting, which I would rather know than have you end the hour flooded and quietly not book again.

Should I expect to feel better afterwards?

Some hours end lighter and some end with more in the room than you brought into it.

The unsettling ones are not failures, and the relieving ones are not evidence. Relief usually comes from being heard, which is real and is not yet change. Discomfort usually comes from having looked directly at something you had been managing with glances, and it is uncomfortable because the thing is now in plain sight. How you feel closing the laptop tells you about that hour, not about the work.

Over a longer stretch, watch for different things: that you can describe your situation more accurately than you could at the start, that something small in ordinary life has moved the way you wanted, that you dread the session less, or that you dread it for a reason you can now name. Treat those as aids to understanding, not as tests. Change comes unevenly, sometimes as a slow loosening and sometimes as a reaction that simply does not arrive any more, and neither of those reads well week by week.

What do I do when what I expected is not there?

Say it in the session, to the therapist, before you decide whether to leave.

This is the hardest thing to ask of yourself, and it decides more outcomes than the choice of therapist did. Stopping early is common. Pooled across 76 trials of one talking therapy and the treatments it was compared with, close to 18 in 100 people ended before the course was finished, a figure from research settings rather than a count of ordinary practice.² What I see is that the people who stop quietly are rarely the ones who were served worst. They are the ones for whom vanishing was easier than objecting, and online it is easier still: there is no reception, no coat, nobody to walk past.

• Name the mismatch. “I expected more questions.” “I left last week feeling worse and said nothing.” A therapist worth keeping treats that as material.
• Ask what they make of it so far. You can ask for their working understanding in plain language, and ask how much of it they still hold open.
• Ask what the work is for. If neither of you can answer, that deserves a session of its own rather than another month of good conversation.
• Review it deliberately. Agree a point at which you both look at how it is going, so the decision gets made on purpose instead of by not rebooking.

Those steps assume a competent therapist who is simply not yet the right one for you. They are not the answer where something has gone wrong: where you leave diminished rather than challenged, where what you said in confidence appears outside the hour, or where your discomfort is about the therapist’s conduct rather than about what you came to work on. There the mismatch is not yours to repair, and ending it is the sound response rather than a failure of nerve. If you want to test any of this before committing, a first conversation about online psychotherapy in English is what it is for.

Questions people ask

Can I ask about a therapist’s training before I book?

Yes, and how the question lands tells you as much as the answer. Ask what they qualified in and how long they have practised; a therapist who hears a straight question as a challenge has shown you something useful.

Do I have to explain everything in the first session?

No, you decide what comes in and when. Saying there is something you are not ready to talk about yet is a genuine contribution rather than an evasion.

How long before I can tell whether it is working?

Long enough to get past the account of the problem and into the detail of it, which for most people is more than one session. A review agreed with the therapist is more use than a fixed number of weeks.

Should I expect advice about what to do?

Expect clear thinking and a view, not instructions about your life. Where a practical step is obviously worth taking I will say so, and the decisions about your relationship or your work stay with you.

Does any of this change when the sessions are online?

What you can expect does not change; what you have to say out loud does. A screen removes the small bodily signals I would otherwise notice, so naming them yourself restores most of what was lost.

Sources

1. Effects of therapeutic alliance on patients with major depressive disorder: a literature review. Frontiers in Psychology, 2025.
2. The Phenomenon of Treatment Dropout, Reasons and Moderators in Acceptance and Commitment Therapy and Other Active Treatments: A Meta-Analytic Review. Clinical Psychology in Europe, 2019.

How this article was written: I choose the subject and the angle. An AI assistant researches and drafts. I read the draft, correct and improve it and stand behind every clinical statement. The professional and clinical judgement is mine.