A substantial share of therapy clients don't complete the exercises their practitioners assign between sessions, and problems with homework are common rather than exceptional (Helbig & Fehm, 2004; Kazantzis et al., 2017). That's not a footnote. That's a structural failure baked into the way most between-session work is designed, delivered, and followed up. And the uncomfortable truth is that the problem almost never sits with the client.
The Homework Compliance Problem No One Wants to Talk About
Let's start with the word itself. "Homework" carries baggage: obligation, drudgery, a power dynamic where one person assigns and the other complies (or doesn't). Most practitioners we speak with know this intuitively. Yet the structure of between-session work in the majority of practices still mirrors the school model: verbal instructions at the end of a session, a worksheet if you're lucky, and a vague hope that the client follows through before next week.
When clients don't complete these exercises, the clinical conversation often shifts to "resistance," "motivation," or "readiness for change." These framings place the burden squarely on the client's shoulders. But look at any other domain of behavioural design, from medication adherence to fitness apps to financial savings tools, and you'll find the same lesson: when large percentages of people fail to do a thing, the problem is almost always the design of the thing, not the people.
This reframing matters clinically, not just philosophically. If non-completion is a design problem, then it's solvable. And solving it has measurable consequences for outcomes, retention, and the sustainability of your practice. The rest of this article is about how to do exactly that.
What the Research Actually Says About Homework and Outcomes
The case for between-session engagement isn't ambiguous. Kazantzis, Whittington, and Dattilio (2010) meta-analysed controlled studies comparing therapy with and without homework and found an effect of d = 0.63 in favour of homework. That's a medium-to-large effect. Clients who engage between sessions simply get better faster.
But there's an important nuance the headline figure obscures. Most studies in this literature measure whether homework was assigned, and sometimes whether it was completed, but very rarely whether it was well designed. Kazantzis and colleagues later acknowledged this gap directly, noting that compliance measures typically treat all homework as equivalent regardless of personalisation, difficulty, or emotional relevance.
This distinction between quantity and quality is critical. A client who completes five generic thought records may show less clinical movement than a client who completes one deeply relevant values-clarification exercise. The evidence supports between-session engagement as a mechanism of change, but it doesn't tell us that any engagement will do. Design quality mediates the effect, even if the research has been slow to measure it.
The quality gap in the evidence base
Mausbach, Moore, Roesch, Cardenas, and Patterson (2010) pooled 23 studies with 2,183 participants and found a small-to-medium association between homework compliance and outcome (r = 0.26). Compliance matters, but it is not the whole story. This points to something practitioners already sense: the process around the exercise matters as much as the exercise itself.
We should be honest about the limits of this evidence. Most homework research comes from CBT protocols with structured, manualised tasks. The evidence base for between-session exercises in ACT, IFS, and other experiential modalities is thinner, though growing. What we can say confidently is that the principle holds across modalities: what clients do between sessions influences outcomes, and how we design that work determines whether they do it at all.
Why Traditional Homework Fails: Five Design Flaws
When we look at how between-session work typically gets assigned across practices, five failure modes emerge repeatedly. These aren't moral failings of busy practitioners. They're structural defaults that persist because no one has redesigned the workflow.
1. Exercises are too generic
A thought record is a thought record is a thought record. Many practitioners reach for the same handful of worksheets regardless of where the client is in their process. A client in early-stage therapy who's still building awareness of emotional patterns receives the same tool as a client deep into committed action. The exercise doesn't meet the client where they are, so it feels irrelevant, and irrelevance is the fastest path to non-completion.
2. Instructions vanish after the session
Here's a scenario most practitioners will recognise: ten minutes before the session ends, you describe an exercise. The client nods. They leave. By the time they're home, the specifics have faded. They remember "something about noticing thoughts," but not the steps, the framing, or the intention behind it. Verbal instructions are the least durable format for delivering complex behavioural guidance. Research on the "forgetting curve" (Murre & Dros, 2015) suggests that without reinforcement, people lose roughly half of newly learned material within an hour, at least for the nonsense syllables used in these experiments.
3. No accountability loop
If a client doesn't complete an exercise, when does the practitioner find out? Typically, at the start of the next session, when the client sheepishly admits they didn't get to it. By then, a week has passed. There's no mid-week nudge, no gentle check-in, no mechanism that makes completion feel expected rather than optional. The absence of a feedback loop doesn't just reduce completion; it signals that the exercise wasn't important enough to follow up on.
4. Exercises feel disconnected from the session
A session might surface deep emotional material: a realisation about a relational pattern, a moment of grief, a breakthrough in self-understanding. Then the practitioner assigns a worksheet that feels like it belongs to a different conversation. The emotional thread that made the session meaningful isn't carried into the between-session work. Clients notice this disconnect, even when they can't articulate it.
5. Assignment without collaboration
Kazantzis, Deane, and Ronan developed a model of therapeutic homework that emphasises collaborative negotiation: the exercise should be co-designed, the client should understand its rationale, and there should be explicit agreement on what "completion" looks like. In practice, time pressure means exercises are often prescribed rather than negotiated. When clients don't feel ownership over the task, they're less likely to prioritise it.
Principles of Effective Between-Session Exercise Design
If those are the failure modes, what does good design look like? Drawing on behavioural design literature, clinical evidence, and what we've observed working with practitioners, five principles emerge.
Micro-dose the effort
Large, open-ended tasks invite procrastination. "Practise mindfulness this week" is not an exercise; it's an aspiration. Effective between-session work is small, specific, and completable in under ten minutes. Fogg (2019) argues that new behaviours are far easier to adopt when the target behaviour is reduced to its smallest viable version. Instead of "journal about your values," try "write one sentence about what mattered most to you today."
Use contextual cues
Linking exercises to existing daily routines dramatically increases follow-through. "After your morning coffee, spend two minutes noticing your emotional weather" leverages a habit stack, a principle well-established in behaviour change research (Lally & Gardner, 2013). The exercise stops being an extra thing to remember and starts being attached to something the client already does.
Negotiate collaboratively
The exercise should feel like the client's idea, or at least like a shared decision. This means discussing the rationale, adjusting the difficulty, and explicitly asking: "Does this feel doable? What might get in the way?" Kazantzis and colleagues' model of homework places this kind of collaborative planning at its centre. It takes an extra three minutes. Those three minutes pay for themselves many times over.
Maintain emotional continuity
The best between-session exercises extend the emotional thread of the session rather than introducing something new. If a session explored the client's relationship with self-criticism, the between-session exercise should directly engage that theme, not pivot to relaxation techniques because they're easier to assign. Emotional resonance is what makes an exercise feel meaningful rather than mechanical.
Build in feedback loops
Clients need to know their engagement is seen. This doesn't require real-time surveillance; it requires a signal that completion matters. A mid-week automated check-in, a brief "how did it go?" prompt, or visible tracking that both practitioner and client can review at the next session. Feedback loops transform between-session work from a solitary obligation into a relational act. For practitioners exploring how to build a data culture around outcome tracking, between-session completion data is a natural starting point.
ACT Between Sessions: Designing Experiential Exercises That Stick
Acceptance and Commitment Therapy lends itself to between-session work better than most modalities, because its core processes are experiential rather than purely cognitive. But that same experiential quality creates a design challenge: ACT exercises often feel meaningful in the room and vague outside it. The key is translating process-level work into concrete, repeatable micro-practices.
Defusion exercises
In session, a practitioner might guide a client through "I'm having the thought that..." or the "leaves on a stream" exercise. Between sessions, this can be distilled into a daily two-minute practice: notice one sticky thought, label it ("There's the 'I'm not good enough' story"), and observe it without engaging. Digital delivery is particularly well-suited here, because a daily prompt at a consistent time can cue the practice. Evidence on digital delivery of ACT interventions suggests that guided, structured digital exercises can replicate meaningful elements of in-session work.
Values clarification
Values work is inherently reflective and translates naturally to journaling prompts and brief self-assessments. A between-session exercise might ask: "Name one action you took today that aligned with a value you care about. Name one you'd do differently." This moves values from an abstract conversation into lived daily review. The specificity matters: "reflect on your values" is too broad; "identify one valued action" is completable.
Committed action
Committed action is where between-session work becomes behavioural. The exercise should name the specific behaviour, the context in which it will happen, and the valued direction it serves. "This week, I will call my sister on Wednesday evening because connection matters to me." This aligns with implementation intention research (Gollwitzer, 1999), which shows that specifying when, where, and how dramatically increases behavioural follow-through.
Levin, Haeger, Pierce, and Twohig found that a web-based ACT self-help programme for college students improved overall distress, anxiety, depression and positive mental health compared with a waitlist, with gains in mindful acceptance and valued living mediating the effects. The evidence base for digitally delivered ACT is growing, and it suggests that well-designed between-session exercises can carry real therapeutic weight. Practitioners interested in tracking psychological flexibility as a clinical KPI will find that between-session ACT exercises generate the data needed to do so.
IFS Between Sessions: Supporting Parts Work Outside the Therapy Room
Internal Family Systems presents a distinct design challenge. Parts work can surface intense material, and both practitioners and clients reasonably worry about doing this work without therapist presence. The solution isn't to avoid between-session IFS work entirely; it's to design it with appropriate structure and guardrails.
Self-led check-ins
The simplest and safest between-session IFS exercise is a daily Self-energy check-in. The client pauses, notices which parts are active, and practises accessing Self, the compassionate, curious internal observer. This doesn't require processing trauma or engaging with burdened parts. It builds the muscle of internal awareness that makes in-session work more efficient. A structured prompt might ask: "Which parts showed up most today? Can you notice them without blending with them?"
Journaling with guardrails
Journaling can support IFS work beautifully, but only with clear boundaries. An effective between-session journaling prompt doesn't invite the client to go deep into exile material unsupported. Instead, it might focus on protector parts: "Notice a moment today when a protector part was active. What was it trying to protect you from? Can you appreciate its intention?" This keeps the work at the protector level, which is generally safe for between-session exploration, while building the client's capacity for the deeper work that happens in session.
Graded exposure to internal experience
For clients further along in their IFS process, between-session exercises can gradually increase in depth. Early exercises focus on noticing and naming parts. Intermediate exercises invite brief, curious contact. Advanced exercises, always negotiated with the practitioner, might involve short, time-bounded check-ins with parts the client has already begun to unburden in session. The key is that the practitioner and client agree on the exercise's scope and the client knows how to regulate if material becomes overwhelming.
For a deeper look at how IFS and digital platforms complement each other, the design principles are consistent: structure, safety, and practitioner oversight. Digital delivery can provide the structured prompts and guardrails that make between-session IFS work feasible without making it reckless.
The Role of Digital Tools in Closing the Completion Gap
Let's return to the five design flaws identified earlier and examine how digital therapeutics, as a category, address each one. This isn't about any single platform. It's about what the technology enables when designed thoughtfully.
Personalisation through adaptive content
A well-designed DTx platform allows practitioners to assign exercises matched to the client's stage of therapy, presenting concern, and modality. Instead of a drawer full of identical worksheets, the practitioner selects from a library of exercises that can be tailored to the individual. Some platforms use adaptive algorithms to suggest exercises based on the client's previous responses; others rely on practitioner curation. Either way, the exercise fits the client rather than demanding the client fit the exercise.
Persistent accessibility
Digital delivery solves the forgetting problem. Instructions don't vanish after the session; they live on the client's phone, accessible when the client is ready to engage. This alone represents a significant improvement over verbal instructions. The client can revisit the rationale, review the steps, and engage at the time that suits their life rather than trying to reconstruct a conversation from memory.
Automated reminders and check-ins
Push notifications and scheduled prompts create the accountability loop that's absent in traditional homework. These nudges don't need to be intrusive; a single, well-timed reminder ("Your two-minute values check-in is ready") can meaningfully increase completion rates. Evidence from medication adherence points the same way: a meta-analysis of 16 trials by Thakkar and colleagues found that text-message reminders roughly doubled the odds of adherence in chronic disease (odds ratio 2.11), though most trials were short and relied on self-report. That's a meaningful lift for a low-cost intervention.
Session-linked exercise assignment
The best digital tools allow practitioners to assign exercises during or immediately after a session, creating a direct link between the session's emotional content and the between-session task. The client receives the exercise while the session is still fresh. This maintains emotional continuity and signals that the exercise isn't generic filler; it's a direct extension of the work they just did together.
Real-time practitioner visibility
Digital platforms can show practitioners which exercises were completed, how clients responded, and where they struggled, all before the next session begins. This transforms the opening of a session from "Did you do the homework?" to "I noticed you found the defusion exercise difficult on Thursday. Tell me about that." The data enables a more responsive, personalised clinical conversation. For practitioners already embracing measurement-based care, between-session engagement data adds a powerful layer of real-time clinical intelligence.
Measuring What Matters: Tracking Completion, Quality, and Clinical Impact
Most practitioners who track between-session engagement at all are measuring a binary: did the client complete the exercise, yes or no? This is better than nothing, but it misses the dimensions that matter most.
Completion rate
Start here, but don't stop here. Track what percentage of assigned exercises are completed across your caseload. If your overall completion rate is below 50%, that's a signal to redesign your exercises before questioning your clients' motivation. Benchmarking this number over time gives you a clear indicator of whether design changes are working.
Engagement quality
Not all completions are equal. A client who spends thirty seconds ticking boxes is engaging differently from one who writes a thoughtful reflection. Where possible, capture qualitative data: free-text responses, self-reported effort, or ratings of how relevant the exercise felt. These data points help you distinguish meaningful engagement from performative compliance.
Self-reported usefulness
A simple post-exercise question, "How useful did you find this exercise?" on a five-point scale, gives you remarkably actionable data. Exercises that consistently score low can be redesigned or replaced. Exercises that score high across clients may be worth making a standard part of your between-session library. This is a lightweight way to run ongoing quality improvement without formal research protocols.
Linking engagement to outcomes
The most sophisticated level of measurement connects between-session engagement patterns to session-over-session outcome change. If you're tracking outcomes using validated measures (the ORS, PHQ-9, AAQ-II, or similar), you can begin to observe whether clients who engage between sessions show steeper improvement trajectories. This won't give you causal proof in a single practice, but it provides clinical signal that helps you refine your approach.
Practitioners who want to build this kind of evidence loop should look at aggregated patterns across their caseload rather than individual cases. Over time, you'll develop a data-informed understanding of which types of between-session exercises drive the most clinical movement for your client population.
A Practical Framework for Redesigning Your Between-Session Strategy
Theory is useful. Implementation is better. Here's a five-step framework you can apply regardless of your modality, client population, or tech stack.
Step 1: Audit your current practices
Spend two weeks simply noticing how you assign between-session exercises. Are you assigning them consistently? Are they personalised or generic? Are you negotiating them collaboratively or prescribing them in the final minutes? Are instructions written or verbal? This isn't about judgement; it's about baseline awareness. Most practitioners are surprised by what they find.
Step 2: Identify where clients drop off
Ask your clients directly: "What happened with the exercise I suggested last week?" Listen for patterns. Do they forget the instructions? Do they start but not finish? Do they find the exercises irrelevant? Do they intend to do them but run out of time? Each pattern points to a different design flaw and, therefore, a different solution.
Step 3: Redesign exercises using the five principles
Take your most commonly assigned exercises and redesign them. Make them smaller (micro-dose the effort). Attach them to daily routines (contextual cues). Negotiate them with the client (collaborative design). Ensure they connect to the session's emotional content (emotional continuity). Build in a check-in mechanism (feedback loops). You don't need to overhaul everything at once. Start with two or three exercises and iterate.
Step 4: Choose a delivery mechanism
This might be digital, but it doesn't have to be. Some practitioners find that sending a brief text message with the exercise instructions is enough. Others use voice memos. Others adopt a dedicated DTx platform. Afterglow, for example, supports ACT- and IFS-informed between-session exercises with practitioner-facing dashboards, but the right tool depends on your practice's needs, budget, and client demographics. The delivery mechanism matters less than the design principles behind it.
Step 5: Track and iterate
Measure completion rates. Ask clients about usefulness. Notice which exercises get done and which don't. Adjust. This isn't a one-time project; it's an ongoing practice. The practitioners who see the best between-session engagement are the ones who treat exercise design as a skill to be refined, not a box to be ticked.
The Future of Between-Session Care
The therapy session has been the atom of clinical work for over a century: the irreducible unit around which everything else orbits. Between-session engagement has been secondary at best, invisible at worst. That hierarchy is shifting.
As blended care models mature, between-session work stops being "homework" and starts being a core clinical intervention. The practitioners and clinics that design this work deliberately, measuring its impact, adapting it to individual clients, and integrating it into their clinical workflow, will see better outcomes. The evidence already supports this. What's changing is the infrastructure to make it practical.
We're also moving toward a world where between-session engagement data informs clinical decision-making in real time. Imagine beginning a session already knowing which exercises your client found challenging, which they found meaningful, and how their self-reported wellbeing shifted across the week. That's not science fiction. It's the direction in which digital therapeutics are already headed.
The question for practitioners isn't whether between-session engagement matters. The evidence settled that years ago. The question is whether you're willing to treat it as a design problem worth solving. The clients who leave your office this week deserve exercises that are built to work, not just assigned to fill the gap.
References
- Fogg, B. J. (2019). Tiny Habits: The Small Changes That Change Everything. Houghton Mifflin Harcourt.
- Gollwitzer, P. M. (1999). Implementation intentions: Strong effects of simple plans. American Psychologist, 54(7), 493–503.
- Helbig, S., & Fehm, L. (2004). Problems with homework in CBT: Rare exception or rather frequent? Behavioural and Cognitive Psychotherapy, 32(3), 291–301.
- Kazantzis, N., Deane, F. P., & Ronan, K. R. (2000). Homework assignments in cognitive and behavioral therapy: A meta-analysis. Clinical Psychology: Science and Practice, 7(2), 189–202.
- Kazantzis, N., Deane, F. P., & Ronan, K. R. (2005). Assessment of homework completion. In N. Kazantzis, F. P. Deane, K. R. Ronan, & L. L'Abate (Eds.), Using Homework Assignments in Cognitive Behavior Therapy (pp. 61–74). Routledge.
- Kazantzis, N., Whittington, C., & Dattilio, F. (2010). Meta-analysis of homework effects in cognitive and behavioral therapy: A replication and extension. Clinical Psychology: Science and Practice, 17(2), 144–156.
- Kazantzis, N., Brownfield, N. R., Mosely, L., Usatoff, A. S., & Flighty, A. J. (2017). Homework in cognitive behavioral therapy: A systematic review of adherence assessment in anxiety and depression (2011–2016). Psychiatric Clinics of North America, 40(4), 625–639.
- Lally, P., & Gardner, B. (2013). Promoting habit formation. Health Psychology Review, 7(sup1), S137–S158.
- Levin, M. E., Haeger, J. A., Pierce, B. G., & Twohig, M. P. (2017). Web-based acceptance and commitment therapy for mental health problems in college students: A randomized controlled trial. Behavior Modification, 41(1), 141–162.
- Mausbach, B. T., Moore, R., Roesch, S., Cardenas, V., & Patterson, T. L. (2010). The relationship between homework compliance and therapy outcomes: An updated meta-analysis. Cognitive Therapy and Research, 34(5), 429–438.
- Thakkar, J., Kurup, R., Laba, T.-L., Santo, K., Thiagalingam, A., Rodgers, A., Woodward, M., Redfern, J., & Chow, C. K. (2016). Mobile telephone text messaging for medication adherence in chronic disease: A meta-analysis. JAMA Internal Medicine, 176(3), 340–349.