S
Supervisia
Sign inStart free →
The CBT career arc — train, supervise, practise, research. One platform, every stage.
General

The CBT career arc — train, supervise, practise, research. One platform, every stage.

The CBT career, taken seriously, runs across four distinct phases — and most clinicians are operating in two or three of them at once. The professional infrastructure that supports each phase was built by people who understood that phase and not the others. The cost of that fragmentation is invisible until you try to move data across the boundary.

9 April 20268 min read

The CBT career, taken seriously, runs through four distinct phases. Trainees learning under supervision. Supervisors holding the next cohort's competence development. Qualified clinicians delivering routine care across a caseload. Researchers running the trials that produce the evidence base the other three are working from.

This is a useful enough framing on its own. What makes it more useful is the recognition that the same clinician usually moves through all four phases across a career — and that at any given point, most CBT therapists are operating in two or three of them simultaneously. The newly qualified therapist is a clinician with a caseload and, increasingly often, a junior trainee they are informally supervising. The experienced supervisor is also a clinician, and may also be a trial therapist on a study running through their service. The researcher is, almost without exception, also a supervisor and often still a clinician. The phases are not stages one leaves behind. They are layers that accumulate.

The infrastructure that supports each phase was, almost without exception, built by people who understood that phase deeply and the other three not at all.

Four phases, four infrastructure stacks, four data layers

Training portals were built by educators. They model the trainee well — curriculum sequencing, competency tracking against a programme's standards, placement records, reflective practice logs. The data belongs to the trainee while they are in the programme and is, in most cases, archived once the programme finishes. The trainee leaves with a certificate and an account they will not log back into.

Supervision tools were built by audit consultants, or by supervisors who built their own spreadsheets and then turned them into products. They model the supervision episode well — session date, content categories, action points, signed-off CPD hours — and they model the supervisor's portfolio assembly badly. They capture the supervision events but rarely capture the supervisee's underlying clinical work that the supervision is supposed to be about. A supervisor reviewing a supervisee's CTS-R competence over time would not, in most existing supervision tools, find that data because the tools were not designed for it.

Electronic medical records were built by insurance companies and, in the UK, by NHS commissioning bodies — primarily to support billing, governance, and outcomes reporting. They model the clinical encounter well enough for those purposes. They model the clinician's developmental record — the sequence of competence development, the supervision that has shaped it, the patterns across a caseload — not at all. The clinician is a name attached to a session, not a developmental subject with a trajectory.

Research dashboards were built by trial methodologists. They model the trial well, and they model the therapist delivering the trial intervention as a fidelity risk to be monitored, episodically, by a separate rater team. The trial-fidelity infrastructure question is taken up in why CBT trials need continuous fidelity infrastructure; for present purposes, the relevant point is that the trial therapist's fidelity data lives entirely inside the trial and is invisible to the same therapist's routine practice record, supervision record, or training history.

Four stacks. Four data layers. None of them speak to each other, and none of them recognise that the user might be wearing multiple hats today, was wearing different hats five years ago, and will be wearing different ones again five years from now.

The continuity problem

The cost of this fragmentation is small in any single moment and large when one tries to do anything that crosses the boundaries.

A trainee's CTS-R record from their training programme does not follow them into qualified practice. The competence development that was carefully tracked over two or three years of training — the items they consistently scored well on, the items where they kept needing supervisory support, the techniques they had begun to find fluent — becomes inaccessible at the point of qualification. The newly qualified clinician walks into their first post with no instrumented record of what they are good at and what they are still developing. Their next supervisor, however well-intentioned, starts from scratch.

A qualified clinician's supervision portfolio does not feed into their BABCP accreditation evidence. The supervision sessions they have attended, the CPD hours they have logged, the case discussions they have brought — all of it sits in one or more supervision tools, in formats that do not map onto the accreditation portfolio's evidence requirements. At portfolio assembly time, the clinician spends weeks reconstructing from emails, calendars, and memory what should have been a continuous record. The supervision contracts problem compounds this — the underlying agreement that governed the supervision is often itself irretrievable.

A research therapist's trial fidelity data is invisible to their routine supervision. The trial rated their sessions, generated CTS-R or modality-specific scores, fed those scores into the trial's fidelity dashboard, and then archived the data inside the trial's data architecture. The same therapist's routine supervision continues alongside, with no awareness of what the trial rating data showed.

A supervisor running a deliberate-practice-informed supervision relationship cannot easily reference the supervisee's drill data, because the drills happened in a training portal that was archived years ago. The deliberate practice evidence suggests that structured, feedback-loaded practice is what differentiates highly effective therapists; the infrastructure that would support it across a career rather than during a training programme has not historically existed.

A clinician trying to monitor their own therapist drift has nowhere to look. Their qualification record sits in one place, their supervision history in another, their outcomes data in a third, and any trial fidelity ratings in a fourth. The pattern across all four would be visible if the data were in one place. In practice it is in four places, and the pattern is invisible.

Each of these is a small inconvenience in any given month. Across a career, they amount to something larger. A profession whose own data is so fragmented that its members cannot easily look back across their own development is a profession whose practitioners are operating with less self-knowledge than the evidence base would justify.

The shared signal that is being fragmented

The underlying observation is that the same kind of data is generated across all four phases. Clinician behaviour, in a session — whether the session is with a real client, a simulated client, a trial participant, or a supervisee discussing a case — can be observed, rated, and recorded against a defined competency framework. The CTS-R is the obvious instrument for CBT, though others exist for other modalities.

That data is generated, in some form, in each phase. The trainee generates it in supervised practice. The supervisor generates it as they review the supervisee's work. The clinician generates it whenever a session is recorded and reviewed, which is rarely but happens. The trial therapist generates it constantly, because that is what the trial's fidelity protocol exists to produce.

All four phases produce data of the same fundamental kind. None of the four phases' standard infrastructure shares it with the others. The therapist whose CTS-R was carefully rated through training, then continuously rated through a trial therapist role, then occasionally rated in routine supervision, has no single instrument that would show the trajectory of those scores across their career. The ratings exist; the trajectory does not, because no system was built to hold it.

This is what makes a unified platform a different proposition from a more convenient one. It is not that having one login is convenient. It is that having one underlying competence record makes possible a set of things that fragmented records cannot do — drift detection across a career, supervision that knows what the supervisee has been drilling, trial-to-routine evidence loops, accreditation portfolios that assemble themselves from data that was being captured anyway, deliberate practice that builds on yesterday's rating rather than starting from a generic curriculum each time. Four separate products with four separate data layers cannot do any of this, however well each product handles its own phase.

Supervisia's position

The argument above is, in fairness, an argument for a thing rather than an empirical claim about a thing that fully exists. Supervisia's position is that the CBT career deserves an infrastructure built around the practitioner's development as the persistent subject, with the four pathways — training, supervision, routine practice, research — as views onto a shared record rather than separate products glued together at the marketing layer.

The four pathways are live. The trainee pathway at /train handles AI clients and CTS-R-scored drills for trainees and qualified therapists doing deliberate practice. The supervisor pathway at /supervision handles the supervision contract, session structure, portfolio assembly, and accreditation evidence capture. The companion pathway at /companion handles the between-session and routine-practice layer that supports the evidence on homework adherence and CBT outcomes. The research pathway at /research handles trial fidelity, rater workflows, and the continuous fidelity infrastructure that the trial literature has been waiting on.

The cross-pillar data layer — the shared underlying competence record that lets a clinician's training scores inform their first qualified supervision, that lets a supervisor see their supervisee's drill data, that lets a trial therapist's fidelity ratings flow into their routine supervision record — is the work that distinguishes the unified-platform position from the four-products-glued-together position. Some of that integration is mature; some of it is in development; some of it is on a roadmap that depends on the field's appetite for unified records in a domain where data has historically been carefully siloed for sound reasons of confidentiality and governance.

The honest framing is this. The idea of the unified platform is what makes Supervisia worth describing as one thing rather than four. The execution of that idea, as of the time of writing, is more complete in some pillars than in others. The argument for the unified platform is not that the integration is finished — it is that the structural problem the integration solves is large enough to be worth solving, and that the existing four-stack arrangement is not going to solve it incrementally on its own.

A practical entry point

This raises an obvious question for anyone reading this. If the cross-pillar integration is the point, but the integration is in progress rather than finished, where does an individual practitioner start?

The honest answer is: at the pathway closest to where you are now. The trainee or qualified therapist doing deliberate practice starts at the training pathway. The BABCP supervisor whose portfolio assembly is approaching starts at the supervision pathway. The private-practice clinician whose between-session structure is the limiting factor starts at the companion pathway. The trial team whose fidelity instrumentation is the bottleneck starts at the research pathway.

Each pathway is useful in its own right. None of them require the others to be present to deliver value. What the unified platform offers is what unlocks over time, as the practitioner's career arc starts to cross the boundaries the four-stack arrangement made invisible — the qualified therapist who becomes a supervisor and finds their training-era CTS-R record still in the system, the supervisor who joins a trial and finds their fidelity ratings flowing into their existing portfolio, the trial therapist who returns to routine practice with the trial's competence record still attached to their professional history.

The CBT career runs through four phases. The profession's infrastructure has not, until recently, been built as though that were the relevant unit of analysis. The argument for treating it as the relevant unit is mostly an argument from the data — from the consistent finding across the drift, deliberate practice, and trial fidelity literatures that what differentiates competent practice from drifting practice, and competent trial delivery from drifting trial delivery, is the continuity of observation and feedback across time. Continuity is hard to build inside any one phase. It is easier to build when the phases share a record.

Supervisia is built around the four pathways and the shared record beneath them.

Pick the pathway closest to where you are now and start there. The trainee or deliberate-practice user starts at Train. The BABCP supervisor starts at Supervision. The private-practice clinician starts at Companion. The trial team starts at Research. The cross-pillar integration is what unlocks over time — as the same practitioner returns wearing a different hat, the underlying competence record is the thing that follows them.

See the full platform overview →

References

  • Chow, D. L., Miller, S. D., Seidel, J. A., Kane, R. T., Thornton, J. A. & Andrews, W. P. (2015). The role of deliberate practice in the development of highly effective psychotherapists. Psychotherapy, 52(3), 337–345. DOI: 10.1037/pst0000015
  • Liness, S. et al. (2019). Clinical supervision in cognitive behavior therapy improves therapists' competence: A single-case experimental pilot study. PubMed: 32213046.
  • Waller, G. & Turner, H. (2016). Therapist drift redux: Why well-meaning clinicians fail to deliver evidence-based therapy, and how to get back on track. Behaviour Research and Therapy, 77, 129–137. DOI: 10.1016/j.brat.2016.01.007. PubMed: 26752326.

Last updated: May 2026

Pick the pathway closest to where you are now

Start free — no card required.

Pick the pathway closest to where you are now →
← Back to home