Context
piar is a psychotherapy practice in Berlin with seven therapists. As Principal Product Designer and Design Strategist (self-employed), I led a 12-month service design engagement to reposition the practice and redesign its patient-facing services. The team consisted of two managing directors and seven therapists who participated in workshops on a rotating basis. Art Direction was handled by Tanja Ritzki. The project operated on a limited budget with no prior experience in service design or product development within the practice.
Starting point
In Germany, patients wait an average of 142 days from first consultation to first therapy session (source: BPtK). Approximately 60% of people with mental health conditions do not seek professional help due to stigma (source: BMG). piar was directly affected by both dynamics.
The practice faced increasing therapy requests based on unrealistic expectations, overloading therapists. Colleagues and patients perceived the practice as outdated. Treatment specialisations were not transparent. No digital infrastructure existed for patient intake, information delivery, or communication.
These problems were connected: unclear information led to mismatched requests, which overloaded therapists, which reduced capacity for patients who did match. The project started without a defined scope — goals, standards, and target group needs had to be discovered during the process.
Approach
Phase 1 — Scope. Briefing interviews and a scoping workshop with the full team defined requirements, constraints, and target groups. An ecosystem map visualised the complexity for the therapists, reducing cognitive load. A scoping document summarised outcomes for management.
Phase 2 — Audit. CompSet analysis of competing psychotherapy websites and digital health platforms using a feature heatmap. Mapping of the existing patient request flow from first contact to follow-up, identifying breakdowns in information access, communication, data handling, and wait time management.
Phase 3 — Explore. Secondary research on psychotherapy access barriers and digital health trends. A behavioural persona workshop with therapists produced behaviour-based profiles — grounded in observed actions rather than demographics — to avoid stereotyping and ensure patient-centred decisions.
Phase 4 — Detect. A patient behaviour journey workshop mapped the full patient path from pre-treatment through post-care, including emotional states, cognitive biases (availability heuristic, confirmation bias, status quo bias), decision points, and intervention opportunities.
Phase 5 — Reframe. An outcome logic map (based on Amy Bucher and Matt Wallaert) structured the relationship between design interventions, behavioural triggers, barriers, and intended outcomes. Baseline measurements were collected over eight weeks by therapists tracking requests and rejections.
Phase 6 — Design. Design principles derived from health facility and emergency design research: rounded corners, soft effects, nature imagery, symmetry, no direct eye contact in photography. WCAG 2.2 compliance. Content rewritten from B2 to B1 language level. Navigation structured using Hick's Law. Three direct entry points for persona-based onboarding. Comparison tables using psychoeducation principles. A digitised request process designed with the Spark Effect pattern. Psychoeducation content integrated into wait time to transform idle time into preparation. Built as a CMS website in Framer.
Decisions
Scope expanded from website to service redesign. The initial brief was a corporate website. During scoping, it became clear that the website could not function without redesigning the underlying service processes. This expanded the project from a digital deliverable to a service design engagement.
Behavioural personas over traditional personas. Demographic personas risk reinforcing stereotypes about mental health patients. Behavioural personas based on observed actions provided more actionable design input and avoided stigmatising assumptions.
Transparency over conversion optimisation. The website visualises therapist workload (125 patients/week, 500 therapy hours/month) and the full therapy timeline. This was a deliberate choice to set realistic expectations and reduce mismatched requests rather than maximise inquiry volume.
Results
Delivered: Scoping document, ecosystem map, CompSet analysis with feature heatmap, patient request flow audit, behavioural personas, patient behaviour journey map, outcome logic map with 8-week baseline measurements, design principles, information architecture, CMS website in Framer with digitised request process, psychoeducation content system, and accessibility-optimised content at B1 language level (evidence: high — artefacts documented). iF Design Award 2024, Finalist.
External validation: iF Design Award 2024 — Preselection Score 313/500 (threshold: 266). Top-scoring criterion: Function (73/100, field average: 54). 10,807 submissions across 72 countries. (evidence: high — scorecard documented).
In progress at project end: Content creation by therapists for diagnosis-specific pages. CMS handover for independent content updates. Whether the therapists maintained and expanded the content after handover is not documented.
Not achieved: Post-launch impact measurement. No data on request volume changes, therapist workload reduction, or patient satisfaction after implementation. Whether the baseline measurements led to measurable change is not recorded. No user testing was conducted within the documented period.
Learnings
Working with a team that had no prior experience in service design required translating every method into accessible formats. The ecosystem map and printed workshop guides replaced facilitator-led sessions with self-guided group work. This built ownership but reduced control over workshop quality. In retrospect, a brief calibration session before each workshop would have reduced variance without removing autonomy.
Dynamic scope adjustment was both necessary and costly. The project scope changed repeatedly as new insights emerged. This produced a more comprehensive solution but created pressure on timeline and budget. Defining explicit scope boundaries earlier — even if provisional — would have made the expansions more visible and negotiable.
Collecting baseline measurements without establishing a post-launch measurement plan was an incomplete decision. The 8-week baseline tracking was methodologically sound, but without a defined follow-up window, the data cannot demonstrate impact. Building the measurement cadence into the project deliverables — not just the research phase — would have closed this gap.