Who Actually Is The Client In Family And Couples Therapy
The question is more loaded than most people think. In individual therapy, the client is pretty straightforward. One person shows up, pays, gets treatment. With couples and families, the structure breaks down immediately. There is no single paying individual whose problems are being solved. At least not in any meaningful way. The client in family and couples therapy is the relationship system itself. Not one person, not even "both people equally" in any simple sense. The treatment target is the pattern of interaction between them, and you as the therapist have to hold that frame consistently or the whole thing collapses into individual therapy by accident. This distinction matters because every ethical and clinical decision branches off of it. Confidentiality limits, session structure, referral decisions, and how you handle disclosure all change depending on who the actual client is. Get this wrong and you will accidentally treat two individual therapy sessions with interruptions, which is worse than doing nothing for most presenting problems.
How This Actually Works In Practice
Start every session by naming the frame out loud. "The work we are doing here is on the system between you two, not on either of you individually." I say this in intake with every new couple. It sounds dry but it prevents an enormous amount of drift. Without that anchor, one partner will inevitably pivot into a solo venting session and the therapist, if not careful, will slip into the familiar solo therapy posture of alliance-building and validation. The alliance in systems work is triangulated. You are allied with the relationship, not with either person. That means you can and will confront both people when they participate in the problematic cycle. You do not let the more articulate or more distressed partner set the narrative agenda. I had a couple once where the wife dominated the first four sessions describing everything the husband did wrong. She was genuinely suffering. But by session five I realized we were running individual therapy dressed in couple's clothing. I stopped it directly and said we would no longer be reviewing complaints one at a time. Instead we mapped the cycle. That shift moved us from twelve sessions to roughly five before they reached a sustainable place.
The Structural Complications
Insurance creates its own set of problems. Most plans require an individual diagnosis code for billing, which means you are literally documenting a pathological label on one person while treating the relationship. This is ethically messy and I have lost count of the times I have sat in supervision wrestling with whether a code like F46.4 was defensible when the problem was clearly interactional. Use it if you have to get paid, but do not confuse the billing code with the clinical reality. Confidentiality is another landmine. You cannot maintain standard individual therapy confidentiality in a couples room. Everything said by either person is known to the other person by definition. I used to try setting complicated rules about what could and could not be shared between sessions. It never worked. People told each other things. The solution is simpler than you might expect. You tell them upfront that anything discussed in session is fair game for both of them to take outside the door, and you build your interventions around that fact rather than fighting it.
Get the Full Details

When The Model Fails You
Family and couples therapy is not appropriate for every situation. Active domestic violence is the clearest contraindication. Bringing two people together when one is using coercion and control is not just ineffective, it is dangerous. I have seen therapists miss this because the presenting complaint looked like a communication problem. The red flags are usually there if you look for them. Isolation from support networks, monitoring of phone and finances, apologies that come only after escalation, and a pattern where one person walks on eggshells. If you suspect this, individual assessment is mandatory before any joint sessions. Do not skip it. Substance use disorders in active phase create another wall. The intoxication or chronic preoccupation with use makes the kind of metacommunication required in systems work virtually impossible. I would rather see a referral to individual addiction treatment first and then revisit the couple's work after sustained recovery. Attempting systems work during active addiction usually just produces one person explaining what the other person does wrong while the other person checks out or becomes defensive.
What Beginners Keep Getting Wrong
The biggest mistake I see is neutrality theater. Therapists think being neutral means giving equal time to both sides and never taking a position. It does not. Systems work requires the therapist to take a position, just not a position with one person against the other. The position is always with the pattern and with change. When you see a cycle where one person pursues and the other withdraws, you name the cycle, you do not pick a side in the cycle. Another failure mode is over-intellectualizing the model. Understanding circular causality on paper is very different from interrupting a circular sequence in real time while two emotionally flooded people are sitting in front of you. You will fumble. You will miss your timing. I still miss cues in live sessions after years of doing this. The way you improve is by getting supervision and by recording your sessions when possible, even if only for your own review. You also need to manage your own urge to rescue. One partner will often come in looking like the victim of the other's behavior. The natural human response is to side with the suffering person. Your job is to resist that response even when it feels morally wrong. The suffering is real, but treating it as individual pathology assigned to one person will not change the relationship. It might temporarily reduce tension in the room, but it will not produce lasting change. That requires targeting the interaction, not the individual.
A Practical Framework For Starting
Session one should cover the structural setup, the definition of the client as the relationship, the limits of confidentiality, and the basic ground rules. You do not need to dive into deep intervention work during the first meeting. Get the frame solid. I usually spend roughly forty-five minutes on structure and the remaining time just observing the natural interaction pattern without trying to change it yet. That observation period tells me more than any questionnaire ever has. Between sessions two and four, begin mapping the presenting problem as a cycle. Use their own language. If they describe a fight about chores, do not reframe it as chores immediately. Track the emotional sequence underneath. What happens when one person asks, what happens when the other does not respond, what happens after that. The content is almost never the actual problem. The sequence is. By session five or six, you should be able to reflect the cycle back to them in a way that neither person disputes. If one of them is still saying that is all the other person's fault, you have not found the cycle yet or the alliance is already broken. Go back to observation. Do not push intervention before the pattern is visible to both of them.

The Hard Parts Nobody Talks About
Some couples do not want systemic work. They want you to fix the other person. You will encounter this frequently and there is no clean way around it. You can refuse to participate in individual blame assignment, which usually forces a conversation about whether therapy is the right fit. I have walked away from cases because the framing was irreconcilable. That is preferable to wasting six months in therapy that is fundamentally misaligned with what the clients actually want. There is also the administrative burden. Documentation in couples and family therapy is heavier than individual work. Consent forms need to address the shared nature of the record. Billing requires careful attention to whether you are running a conjoint session or an individual adjunct session, because the codes and justifications are different. You need a clear policy on what happens if one person wants to stop therapy while the other wants to continue. Some therapists will offer individual sessions within the same treatment, others will refer the continuing person out. Both approaches have merit, but you need to decide before you get asked, not during a crisis. The model works when the relationship itself is the thing worth saving. It does not work when one person is already grieved and gone, or when abuse is present, or when the motivation is entirely external pressure from family or court. Knowing which situation you are in before you commit to the framework saves time and prevents collateral damage. Most of the time, if you get the initial assessment right, the rest of the work follows a relatively clear path.