← Learn · Therapy Progress · 12 min read

How to Know If Therapy Is Working: A Real Method

Most people answer this with vibes. There's a better way: compare what you talked about three months ago to today. Here's how.

Most people try to assess therapy with vibes. They replay last week's session, scan their mood that morning, and ask whether they feel better. That doesn't work, because how you feel about therapy fluctuates session to session. The honest test is comparative. Look at what you were processing three months ago and compare it to what you're processing now. Are the same themes still dominating sessions? Has your language about a specific person, fear, or trigger shifted? Do situations that used to send you spiraling barely come up anymore? That's the evidence. Progress is almost never a feeling. It's a difference between two points in time. The rest of this guide shows you how to actually find that difference instead of ruminating on it.

Why "do I feel better?" is the wrong question

Open any other article on this topic and you'll get a tidy list: you feel calmer, your sleep is better, your friends notice a difference, you handle conflict more gracefully. None of those signs are wrong. They're just not verifiable. You can't audit them against anything, which means what you're calling assessment is actually rumination in a trench coat.

Therapy outcomes are surprisingly hard to feel from the inside. Without a structured way to track progress, both clients and therapists tend to overestimate or underestimate change. Roughly 75% of people who enter psychotherapy show some benefit from it, and about half see meaningful symptom improvement within eight sessions. But most people who quit therapy don't quit because therapy stopped working. They quit because they couldn't tell whether it was working. A widely cited meta-analysis of 669 studies found a premature dropout rate of about 19.7% across adult psychotherapy, with poor perceived progress as a leading driver. If you want the raw numbers — what percentage of people benefit, how long therapy takes to work, dropout rates — see our therapy statistics roundup.

The takeaway isn't "trust the process." It's: use a method that doesn't rely on your mood the day you ask the question.

Evidence over rumination: the actual method

There are two ways to think about whether therapy is working.

Rumination-based

You scan inward. "Do I feel different? Am I happier? Has it been worth it?" The answer changes depending on whether you slept well, whether your week was hard, and whether your last session happened to crack something open or feel flat.

Evidence-based

You look outward at a record. "What was I bringing into sessions in February? What am I bringing now? Has the texture of what I'm working on changed?" The answer doesn't move with your mood, because it's based on what actually happened.

The first method punishes you for having a normal human nervous system. The second one gives you something you can stand on.

The reason this works is that meaningful therapeutic progress almost always shows up as change in content before it shows up as change in feeling. You stop bringing the same fight with your mother to every session. That's progress, even if you don't feel healed. Your language about your boss shifts from "he's destroying me" to "he's a difficult person I have to manage." That's progress, even if work still sucks. You used to spend twenty minutes per session on a specific fear and now it doesn't come up. That's progress, even if you don't feel particularly transformed.

You can only see those shifts by comparing. And you can only compare if you have a record.

Signs other articles list, and the versions you can actually verify

Here's the difference between the formulaic version of this section and a useful one.

What other articles say

Your emotional regulation has improved. Friends and family notice a change. You sleep better and feel calmer. You handle conflict more gracefully. You feel more like yourself.

What you can verify with a record

The same trigger appeared in three old sessions with intense language and barely comes up in the last three. A relationship you used to vent about isn't in the last month of notes. "Anxiety" used to be the dominant theme; now it's something more specific and actionable. You can find the moment your language about a person shifted from blame to ambivalence.

Notice the pattern. The left column is feelings about progress. The right column is artifacts of progress. The right column is something you can show a friend, a therapist, or your future self, and they'll see it too. The left column is something you have to take on faith from a version of yourself that may or may not be having a good day.

Looking for proof vs. looking for continuity

There's a related trap worth flagging because almost everyone falls into it: confusing proof with continuity.

Proof

What you go looking for when you've decided in advance that therapy either is or isn't working. You scan for evidence to confirm the verdict you already have. It's why some people leave a great session convinced therapy is changing their life, and the same people leave a flat session two weeks later convinced it's been a waste.

Continuity

What you go looking for when you're trying to find the thread. You're not asking "is this working?" You're asking "what story is being told across these sessions, and is it moving?"

Progress is almost never a single proof point. It's a thread that runs through twelve weeks of notes and points in a direction. If you only look for proof, you'll miss the thread. If you look for continuity, the proof shows up on its own.

How to do this manually (no app required)

You don't need software to do evidence-based progress tracking. You just need a habit.

After each session, write three things down The main theme. A new insight or reframe. An open question or homework you're carrying into next time. Five minutes max. Don't write a transcript. Write the shape of the session.
Every four to six weeks, reread the last set Don't read them in order looking for highlights. Read them looking for repetition, drift, and disappearance. What themes repeat? What has quietly stopped showing up? What words or framings have changed?
Write a one-paragraph summary of that window "Between weeks 8 and 14, the dominant theme was X. Y came up twice and then disappeared. My language about Z shifted from A to B."
Every three months, compare windows Put the paragraph from months 1 to 3 next to the paragraph from months 4 to 6. The delta between them is the answer to "is this working?"

The manual version of this works, but it has two failure points. Most people don't actually write session notes consistently. And even when they do, most people don't reread them with the right frame.

How Undertone automates the same loop

Undertone handles both failure points. After each session, Undertone's Session Summary captures what came up, what shifted, and what felt unresolved. Three layers of detail, skimmable first, deeper if you want it. You don't have to remember to write anything down. The summary is waiting when you're ready, whether that's the same evening or three days later when your nervous system has settled.

The Pre-Session Brief is what closes the loop. Before your next session, it surfaces what was explored last time, what felt unresolved, what you said you'd try, and anything you tagged in Bring to Next Session during the week. You walk in with continuity instead of walking in cold and spending the first ten minutes trying to remember where you left off.

Over time, the summaries compound. The record itself becomes the comparison tool. When you ask "is therapy working?" you have material to look at, not just a feeling to interrogate. Undertone stores structured outputs only. The Session Summary, Moments, and Pre-Session Brief. Not raw audio. Not transcripts. The point is continuity, not surveillance.

In session: Something lands. You feel it. You'd swear you'll remember it.
After session: The Session Summary captures it in a form you can come back to. Skimmable now, deeper when you have time.
Between sessions: Moments resurface short reflections from what came up. You're not re-reading the whole session. You're meeting one piece of it at a time.
Before next session: The Pre-Session Brief pulls forward themes, unresolved threads, and what you flagged. You walk in already in the conversation.

A reflection check-in you can run right now

If you don't yet have a record to look back on, you can still get partway there with a structured reflection. Set a timer for fifteen minutes and answer these in writing.

Fifteen-minute progress reflection

What's a sentence I would have said about my life three months ago that I would not say today? Be specific. "I was anxious then and I'm anxious now" doesn't count. "I would have said my brother was the problem; I wouldn't say that now" counts.
What's a topic that used to take up real session time that hasn't come up in the last month?
What's a feeling or reaction I used to have automatically that now has a pause in it? Even a one-second pause is a pause.
What's something I now know about my own pattern that I didn't know in week one?
What's still exactly the same as when I started, and is that because it doesn't matter, or because we haven't gotten there yet?

If you can answer at least three of those with something specific, therapy is doing work, even if it doesn't feel dramatic. If you can't answer any of them specifically, that's not a verdict that therapy isn't working. It's a signal that you need a record before you can fairly assess.

Questions to ask your therapist

Bringing this conversation to your therapist directly is one of the most useful things you can do. Good therapists welcome it.

"Can we spend ten minutes reviewing where we started and where we are now?"
"What do you think has shifted in the last few months, and what do you think hasn't?"
"What would you expect to see change in the next eight weeks if this is working?"
"Are there specific patterns you've noticed me repeating that I haven't seen yet?"
"If we wanted to be more deliberate about tracking progress, what would you suggest?"

Notice that none of these questions are "am I getting better?" That question puts your therapist in an awkward position. They can't give you an honest answer in real time because they're inside the work with you. The questions above ask them to do something they can actually do: reflect on the arc, name patterns, and set expectations.

When to give it more time vs. when to consider a change

Evidence-based assessment also helps with the harder question, which is whether to stay with your current therapist.

Stay the course When the themes are shifting but slowly, when sessions feel uncomfortable but productive, or when you can name at least one or two things that have changed even if the bigger picture hasn't.
Raise it directly with your therapist When the same themes have been recycling unchanged for several months, when sessions feel pleasant but not generative, or when you keep leaving with the same homework you came in with.
Consider a different therapist or modality When, after raising it directly, nothing changes. When the relationship feels safe but not challenging enough, or challenging in ways that don't lead anywhere. About one in five clients ends therapy prematurely; many of them would have been better served by switching rather than quitting.

The point isn't that therapy that feels slow isn't working. The point is that you should be able to name what's happening, not guess.

FAQ

How long should it take before I know if therapy is working?

Most people need at least eight to twelve sessions before there's enough material to compare. Research suggests roughly half of clients show measurable improvement within eight sessions, but the experience of recognizing progress lags behind the progress itself by a few weeks, especially without notes.

What if I feel worse since starting therapy?

Feeling worse in the short term is common and often a sign that real material is surfacing rather than that therapy is failing. The check is whether the difficulty has shape and direction. You're working on something specific that's hard, versus feeling diffusely worse with no thread.

Should I tell my therapist I'm not sure if it's working?

Yes. This is one of the highest-leverage conversations you can have. A therapist who responds defensively gives you useful information. A therapist who welcomes it and helps you think about it gives you even more.

Can I track progress without writing things down?

You can, but you'll be less accurate. Memory rewrites session content constantly, especially for emotionally charged material. Even rough notes, three sentences after each session, outperform recall by a wide margin.

Is it normal to question whether therapy is working?

Yes. The people who never question it are usually the ones who quit the soonest, because when doubt eventually arrives, they have no framework for handling it. Structured reflection is what makes the doubt productive instead of corrosive.

If you want the comparative view without building the habit from scratch, start a free Undertone trial and have your session themes tracked automatically.

Get Undertone