Ask around and you will find them everywhere: people who have been in therapy for four, six, nine years. Same therapist, same Tuesday slot, no particular destination. Ask how it's going and you get "it's good to have the space."

Sometimes that is exactly right. Some conditions need long-term care, and some people knowingly buy ongoing support the way you keep a gym membership. No argument.

But a large share of those multi-year therapies are something else: a drift. Nobody set a goal, nobody measures anything, nobody has said the word "discharge" since intake. And here is the uncomfortable part: the drift is not a bug in the system. Structurally speaking, it is the product the system is built to sell.

Follow the money for one minute

Nearly everyone in mental health care gets paid the same way: per session, forever.

The therapist bills a session. The insurer reimburses a session, usually at a discounted panel rate, which means a panel therapist covers rent through volume: a full calendar, indefinitely. The platform, if there is one, takes its cut per session. Line the incentives up and look at them:

• Nobody in the chain earns more when you get better faster.
• Nobody earns less when you plateau for three years.
• A client who stays stuck but keeps attending is, on paper, identical to one who is recovering, and financially better, because they stay longer.

Insurance reimburses visits, not recoveries. There is no CPT code for "client no longer needs me." I am not accusing your therapist of running a con; the overwhelming majority of clinicians are decent people doing hard work. The point is colder than an accusation: when a system pays for attendance and never checks outcomes, drift does not need villains. It happens by default, the way water finds the low spot.

The evidence that this isn't how it has to be

Here is what makes the drift genuinely tragic rather than just expensive: therapy works, and works fast, when it is run like it means it.

Decades of outcome research on structured, evidence-based treatments (CBT, behavioral activation, exposure-based therapies, EFT for couples, and others) find substantial improvement for most responders within roughly 8 to 20 sessions for many common conditions. The dose-response literature in psychotherapy consistently shows the steepest gains happening early, with diminishing returns as unstructured therapy stretches on. And routine outcome monitoring, the simple practice of measuring symptoms regularly and adjusting when the data flatlines, improves outcomes and catches failing treatments that clinician judgment alone misses.

So the field knows how to run therapy that ends. Measured, planned, goal-directed treatment is not a fringe idea; it is what the evidence base has been recommending for decades. It is just not what the payment structure rewards, so it is not what the average consumer experience delivers.

The mediocre-fit trap

Drift has a partner: the so-so match nobody escapes.

Finding a therapist is exhausting, so once you have one, inertia is enormous. And the system provides no checkpoint at which anyone asks whether this particular pairing is working. The insurer doesn't ask. The directory that surfaced the name doesn't follow up: Psychology Today verifies licenses, not the self-typed specialty tags, and won't say how it orders its listings; Zocdoc sells labeled sponsored slots and ranks by who has openings; BetterHelp's own FAQ says your assigned therapist depends partly on availability. None of the big directories collect outcome data on the matches they made. Your therapist is the only professional in the room, and asking them "is this working?" feels like insulting your barber mid-haircut.

So people stay years in therapy that is fine. Pleasant, validating, covered by insurance, and not producing change. The research on therapist effects is blunt: therapists differ meaningfully in outcomes, and fit with the specific problem matters. Staying with a mediocre fit is not neutral; it costs the exact years the right treatment would have used.

Two questions cut through it, wherever you are in therapy right now: What are we working toward, and how will we know? A good therapist loves those questions. If they produce visible discomfort and no answer, you have learned something important.

What therapy built to end looks like

This is how I run my practice, and what I would tell you to demand from any clinician, in-network or out:

1. Diagnostic clarity early. By the end of the first session or two, a plain-language explanation of what is going on. Not a vibe, a formulation.
2. A written plan with a direction. Which approach, targeting what, with a rough expected arc. "We'll see where it goes" is not a plan; it is a subscription.
3. Progress measured out loud. Symptom measures and honest check-ins on a schedule. If the line is flat, the plan changes, not the appointment time.
4. Discharge as the goal from day one. The end is designed in: skills transferred, gains consolidated, door left open. When you have what you came for, we are done.
5. Honesty about the exceptions. Some conditions and some seasons of life genuinely need long-term or ongoing care, and a straight clinician tells you that to your face, with reasons, rather than letting the calendar decide.

Run the math on what this means financially. Sixteen effective sessions at my full $200 rate is $3,200, minus what a PPO plan reimburses out-of-network, often half or more. Four years of weekly $40 copays is over $8,000, on top of premiums, for therapy that may be going nowhere, before counting what four stuck years cost you in work, relationships, and compound misery. The expensive therapy is not the one with the bigger sticker. It is the one that doesn't end.

FAQ: therapy length, cost, and getting unstuck

How long should therapy take? For many common presentations, evidence-based protocols run roughly 8 to 20 sessions, with early gains visible in the first month. Deeper or layered problems take longer, legitimately. The honest answer is a range plus a plan for checking progress, never a shrug.

Is long-term therapy always a scam, then? No. Personality-level patterns, complex trauma, chronic conditions, and maintenance care can all justify long-term work, and some people rationally choose ongoing support. The test is not duration; it is whether anyone can say what the work is currently for and point to evidence it is doing that.

How do I know if my current therapy is working? Symptoms moving, behavior changing, people around you noticing, and you doing things you couldn't do before. If you cannot name a change in the last three months, ask the two questions: what are we working toward, and how will we know? The conversation that follows is diagnostic either way.

Isn't "feeling supported" enough of a result? Support matters, and for some seasons it is the point. But if support is the only product after months of paying, you deserve to have chosen that explicitly, not drifted into it. Rented friendship at insurance rates should at least be a decision.

Won't my therapist be offended if I ask about progress? A good one will be pleased; outcome-monitoring research is on your side, and clinicians who measure do better. Visible offense at accountability is information about fit.

How do I leave a therapist who isn't helping without feeling awful? Say it plainly: "I'm not seeing the change I need, and I'm going to try a different approach." You do not owe years of loyalty for months of pleasantness. Any decent clinician will take it professionally, and many will help with the referral.

Why would a therapist run their practice in a way that ends treatment faster? Isn't that against their own interest? Short-term yes, long-term no. A practice built on results runs on referrals and reputation instead of retention, clients who finish send their friends. It also happens to be the ethical direction the field's own guidelines point: effective treatment, no longer than necessary.

Does paying out-of-network actually change the incentives? Partly. It frees the clinician from panel-rate volume math and frees you from network-limited choices, which widens the fit pool. It does not automatically buy you goal-directed care, you still have to demand the plan, but it removes the structure that most rewards drift. The insurance guide covers how reimbursement works.

What if I try focused therapy and it turns out my problem needs longer care? Then you will know that from data instead of drift, and the longer care will have a rationale, a target, and checkpoints. That is a completely different thing than year four of "it's good to have the space."

What does this look like with Dr. Chan concretely? Free 15-minute consultation. First session or two: goals, history, diagnostic clarity, written plan. Then active weekly work with skills between sessions and progress reviewed openly. When you have what you came for, we end, and the door stays open. $200 per session, insurance accepted through Headway, superbills for out-of-network reimbursement.

The system won't ask. You can.

Nobody in the payment chain is going to ask whether your therapy is working. The system is not built to ask. Which means the question belongs to you, and it is worth asking this week, whether you ask it of your current therapist or a new one.

I'm Dr. Anders Chan, a licensed psychologist in New York. My practice is built to end: diagnosis, plan, measured progress, discharge as the goal. If I'm not your fit, the matching tool will look for someone who is.

This article is educational and is not a diagnosis or a substitute for care from a licensed professional. Treatment length varies by person and condition; nothing here guarantees an outcome. If you are in crisis, call or text 988 (Suicide and Crisis Lifeline) in the US.