The therapy menu is a lie.

The therapy menu is built around the belief that the main task is choosing the right branded method. CBT. EMDR. IFS. ACT. EFT. Twenty-nine names, each with a founder, a theory, and someone willing to explain why it matters.

A reasonable person sees that list and starts optimizing. They google. They rank. They wonder if choosing CBT when they needed psychodynamic work will cost them years.

The field's strangest finding cuts against that fear. When competent active therapies are compared against each other for the problems therapists see most often, the differences are usually small. The human factors are doing more work than the menu admits.

OCD is the clean example. The obsession triggers anxiety, the compulsion relieves it, and the brain learns that the compulsion kept everyone alive. ERP is built to break that loop. Generic warmth does not do the same job.

PTSD has its own version of this. Trauma-focused protocols update memory and meaning in ways generic supportive work usually does less directly. Severe BPD and chronic self-harm are another case: DBT works because the scaffold keeps treatment from collapsing.

That is the pattern. Technique matters when the condition has a known maintenance loop and a protocol built to hit it. For most depression, grief, adjustment, identity work, and diffuse anxiety, the therapist and the fit carry more of the load.

A flat list of therapies is a trap. The mental model begins when the list becomes seven families: cognitive-behavioral, trauma-focused, insight, humanistic, systems, process-specific, and identity-context work.

Once you know the family, you know the therapy's default bet about change. Does it change behavior, update trauma memory, repair attachment, build meaning, shift a couple's cycle, resolve ambivalence, or protect identity in the room?

CBT is not merely evidence-based. It is evidence-saturated. Brainspotting is not merely newer. It has a much smaller evidence base. EMDR is well-supported for PTSD and much thinner outside trauma. Gottman has powerful prediction science and more modest treatment-outcome science.

That matters because marketing turns evidence into a yes-or-no sticker. The useful question is sharper: evidence for what condition, compared with what, by whose lab, in which population?

Some people think their way through pain. Some people feel their way through it. Some want structure, homework, and a sense of forward motion. Others experience structure as cold, and they need an open room before they can tell the truth.

These are clinical variables. A treatment the client believes in tends to get more engagement, less dropout, and more movement. A perfect protocol that the person will not do loses to a solid therapy they can actually stay with.

Identity belongs here too. Faith, culture, language, gender, race, immigration, and community can be the material of therapy rather than background detail. Fit means the person can bring the world they actually live in.

If there is immediate danger, stabilize first. If the concern is OCD, PTSD, severe BPD or chronic self-harm, an eating disorder, substance use with ambivalence, or couples distress, look for the matched specialist protocol.

For most other concerns, choose a therapist you can actually talk to, whose way of working matches how you change, whose constraints fit your life, and whose room can hold the parts of your identity that matter.

Then give it eight to twelve weeks. If the alliance is solid and something is shifting, stay. If nothing is moving, name it in the room. If the alliance never forms, leave. The dodo bird was mostly right. Everybody has won. The question worth obsessing over is which therapist, in this week of your life, can actually be in the room with you.