"Did you get a chance to do the thing we talked about?" You already know the answer from the face. The thought record is blank, the worksheet's in a drawer, and now you're both spending the first ten minutes managing the small shame of undone homework instead of doing the work.
Most of us have quietly made our peace with low homework completion. But the completion rate isn't a fact about your clients' motivation — it's mostly a fact about how the homework was designed. Here's the field-guide version of what changes when you treat it that way.
The adherence problem isn't a motivation problem
Start by retiring the explanation we reach for first. When homework doesn't get done we tend to file it under resistance, ambivalence, or "not ready" — clinical-sounding versions of "the client didn't try hard enough." Sometimes that's real and clinically meaningful. Far more often, the client wanted to and the task was simply built in a way that couldn't survive their week.
That reframe matters because it points at something you can fix. You can't will a client into more motivation from outside. You can absolutely redesign the task. So the useful question isn't "how do I get them to comply," it's "what made this impossible to do, and what would make it doable."
Why most homework doesn't get done
Four design flaws account for most of it. They tend to stack.
- It's in your language, not theirs. A worksheet is a clinician's artifact, organized by your model. When the client is activated, your CBT thought record is the last thing that feels like theirs to reach for.
- It asks for the wrong state. Homework usually requires the calm, reflective state — which is exactly the state that's offline when the pattern actually fires. This is the same state-access problem behind why insight doesn't stick.
- It points away from you. A form the client fills out alone, to bring back for grading, quietly re-enacts disconnection. It's work done at the client, not with them.
- It's scheduled, not situational. "Do this Wednesday" rarely lines up with Thursday at 11pm, when the part actually takes over and the work would matter.
None of these is a character flaw in the client. They're all things you can change before they leave the room.
What makes homework get done
Flip each flaw and you get the four traits of between-session work that survives the week.
- Co-built in the client's own language. Their word for the feeling, their image, their phrase. If they can't say it back to you in their own words, they won't find it midweek.
- Shrunk to one action. Not a practice, not a log — one small thing they could do even while activated. One breath. One sentence. One noticing. Small enough that doing it is almost easier than not.
- Anchored to a moment. Tied to the real situation where the pattern shows up, not to a day on the calendar. The cue does the remembering so the client doesn't have to.
- Routed back. The point isn't completion for its own sake; it's that something comes back into the room — what happened, or that nothing did and why. That makes it relational instead of a test.
A redesign you can run in the last five minutes
This is the part you can use tomorrow. In the final few minutes of session, instead of assigning, build it together:
- Pick the moment, not the day. Name the real situation — the tight chest before the meeting, the urge to reread the text — where the pattern lands.
- Shrink it to one action. Cut the task to the smallest real thing they could do while activated. If it needs the calm state to complete, it's still too big.
- Put it in their words. Have them say it back and write down their version, not yours.
- Set the carry-back. Agree on the one thing they'll bring next session — including "I noticed I couldn't do it, and here's what got in the way," which is genuinely useful clinical material rather than a failure.
What it looks like across modalities
The shape is modality-agnostic; the content isn't.
- CBT. Instead of a full thought record, one catch-and-name in the client's phrase — "there's the 'I'll be exposed' story again" — noticed in the moment, written down later in one line.
- IFS. Not a parts-mapping worksheet but a single noticing: "the foreman is online." The full version of this is its own piece on IFS between sessions.
- Attachment. Rather than "journal about your patterns," one prepared sentence the client can actually send when the urge to withdraw or pursue hits — a small, real reach built in their voice.
- Somatic. Not a body-scan series but one orienting cue the body already knows, available before the thinking comes back online.
When no homework is the right call
Worth saying, because the field over-prescribes: sometimes the right between-session task is none. If the work is genuinely in-session, assigning something to look productive just manufactures another occasion for shame. "Nothing this week — let's let last session settle" is a legitimate clinical choice, and it protects your credibility for the times you do ask for something.
The line — a tool that helps homework get done, not one that replaces you
If you reach for technology to close the gap, hold the category line. A tool that helps the client notice and carry the work back to you — a mirror, in their language, that keeps you in the loop — is one thing. A chatbot that becomes the client's between-session "therapist" is another, riskier thing entirely; general-purpose AI chatbots respond appropriately to distress below 60% of the time against around 93% for licensed clinicians, and anything touching client material has to be HIPAA-compliant by design. I draw that line in AI as clinical tool, not replacement, and the larger frame is in the pillar on keeping clients engaged between sessions.
FAQ
Why don't clients do their therapy homework?
Usually not because they're unmotivated. Most homework fails by design: it's written in the clinician's language, it assumes the calm, reflective state that's offline when the pattern actually fires, it points the client away from the relationship, and it's scheduled for a day rather than tied to the real moment. Fix the design and adherence climbs without any extra pressure on the client.
How do I make therapy homework more effective?
Co-build it in the client's own language, shrink it to one small action they could do even while activated, anchor it to a specific moment rather than a calendar day, and agree on one thing to carry back to session. Effective homework is small, situational, in their words, and routed back to you — not comprehensive, scheduled, in your vocabulary, and graded.
Should I stop assigning homework if clients don't do it?
Stop assigning homework that's designed to fail, not homework itself. Sometimes the right answer is no task at all — the work is in-session and assigning something just to look productive adds shame when it isn't done. When between-session work does help, redesign it so it fits the activated state and the client's language rather than abandoning the idea.
Is between-session homework the same as client engagement?
No. Homework is a task the client completes or doesn't; engagement is whether the therapeutic work stays available to them when the pattern shows up midweek. A client can finish every worksheet and change nothing, or do no formal homework and still catch the pattern in real time and bring it back. The second is engagement — and it's the actual goal.