In the first three sessions, a couples therapist is watching the process far more than the content. They are tracking who speaks first, how each partner reacts while the other is talking, whether either of you can absorb a criticism without immediately defending, and how quickly the temperature rises when a real disagreement surfaces. The specific complaints matter much less than the shape of the exchange.

Most therapists also form a working hypothesis in those first hours, usually about the cycle you are stuck in rather than about who is right. That hypothesis gets tested and revised, but it typically arrives early, because the pattern that runs your relationship at home reproduces itself in the room within about twenty minutes of the first session.

The Very First Thing They Are Watching

Who talks first, and what the other person does while it happens. If one partner delivers a fluent five-minute account of the problem while the other stares at the carpet, the therapist has learned something about the distribution of speech and probably about the distribution of power in the relationship.

They watch for eye rolls, sighs, and small facial dismissals during the other person’s turn. The Gottman Institute categorises exactly these behaviours- sarcasm, eye-rolling, sneering and mockery, as contempt, and describes it as the single greatest predictor of divorce among the four patterns it tracks. It is visible in seconds. A single eye roll means nothing. A pattern of them across three sessions is a different matter entirely.

They also watch what happens to repair attempts. Someone makes a joke to break tension, or reaches out, or softens their tone mid-sentence. What matters is whether the other partner accepts the offer. Couples who repair successfully can argue badly and still be fine. Couples who let repair attempts drop tend to escalate no matter how reasonable the initial topic was.

Physical arrangement gets noticed too, though it is a weaker signal than people assume. Where you sit, whether you angle toward each other, whether you use each other’s names or default to “he” and “she” while your partner is in the room.

The Cycle They Are Mapping Out

By session two, most therapists have sketched a loop. The overwhelmingly common one is pursue and withdraw, sometimes called demand and withdraw in the research, where one partner escalates toward connection through questions, complaints, and volume, while the other shuts down, goes quiet, or leaves the room. Each behavior triggers the other, and both people experience themselves as reacting rather than starting.

What the therapist is trying to identify is the moment the cycle begins, which is almost never where the couple thinks it is. Partners point to the shouting or the silent treatment. The therapist looks earlier, for the small flicker of hurt or fear that preceded either one, since that is the point where intervention actually works.

They are also listening for which of the four corrosive patterns are present: criticism (attacking character rather than behavior), defensiveness, contempt, and stonewalling. Most distressed couples have two of these. The particular combination shapes the treatment plan more than the presenting complaint does.

Naming the cycle out loud is often the first visible deliverable of couples therapy work, and it lands differently than either partner expects, because it removes the question of who started it without excusing anyone’s behavior.

A separate thing they are quietly assessing is whether each partner can take any responsibility at all. A person who can say “I know I do this thing and it makes it worse” in session two is a much better prognostic sign than a person with a perfectly documented case against their spouse.

What They Learn From How Each Partner Answers Questions

Therapists often ask how you met and what the early years were like. This looks like small talk. It is not. Research on how couples narrate their own history has found that the tone of the retelling, whether it is fond and detailed or flat and dismissive, correlates meaningfully with where the relationship is heading. Two people who light up describing a terrible first apartment have something to work with.

They notice who fills silence. They notice whether one partner routinely answers questions addressed to the other, and whether that partner seems relieved or annoyed by it. They notice the difference between someone who cannot access their feelings and someone who is protecting themselves from expressing them, which look identical on the surface and require completely different approaches.

Language gets tracked closely. “We” versus “I” versus “you always.” Absolutes. Whether complaints come with a specific example or arrive as global character verdicts. Whether either partner uses the word divorce, and in what register: as a threat, as a fear, or as a fact they have already accepted privately.

They also register asymmetry in motivation. One partner booked the appointment, chose the therapist, and has read three books. The other showed up. That gap is common and workable, but it is a live issue that a good clinician will name rather than pretend not to see, often within the first two sessions.

The Practical Assessment Running Underneath

Alongside all this, there is a structured assessment happening. Most therapists take individual sessions in weeks two or three, seeing each partner alone for a full hour. That is standard practice, not a sign that things are dire, and it exists to surface anything that will not be said in front of a spouse.

Those solo sessions cover safety, substance use, untreated depression or anxiety, and affairs. Therapists differ on how they handle secrets, and reputable ones state their policy upfront, usually some version of no-secrets, meaning they will not carry a hidden affair indefinitely. Ask about this before you start rather than after, because the policy varies between practices and it materially changes what you should say.

Practically, expect 50 to 90 minute sessions, weekly at first, and a fee commonly in the $175 to $350 range per session privately in large US metros. Many couples therapists ask you to complete an intake questionnaire before session one, which can run 30 to 100 items and covers satisfaction, conflict frequency, intimacy, and shared goals. Insurance coverage is inconsistent for couples work, since it is often billed differently from individual therapy, so verify before the third session rather than after the tenth.

By the end of session three, most therapists will offer you something concrete: a name for the cycle, a rough plan, and an estimate of length. Eight to twenty sessions is the usual frame for focused work. Anyone giving you a guarantee is overselling, and anyone still gathering history in session six without offering a formulation is worth asking directly what they are seeing.

The thing worth doing on your side after those three sessions is a private check. Did the therapist manage the room, meaning did they interrupt escalation rather than watching it happen, and did both of you leave feeling equally heard, including the partner who came reluctantly. A slight imbalance is normal early. A consistent sense that the therapist has a favorite is a reason to raise it in session four, out loud, since a competent clinician will take that feedback seriously rather than defend against it.

Also notice what happened at home between sessions. The most useful early sign is not that you argued less. It is that during an argument, one of you recognized the pattern while it was still running, and said so. That single moment of shared observation is the actual mechanism of change, and it usually shows up long before the arguments stop.