Why finding a therapist is nothing like finding a surgeon

AUTHOR
Prayat Shah
DATE
November 17, 2026
CATEGORY
Stories
Table of contents
SHARE
Cover 100% of your employees, from Day 1.
Woman and man smiling and shaking hands indoors in front of a window with green plants outside.

Finding a therapist feels like it should be as straightforward as finding an orthopaedic surgeon. You break your arm, you look at who's local and well-reviewed, you pick a doctor with the right credentials, you get it set. Credentials map onto outcomes, and everyone with a comparable fracture ends up with comparable care, give or take the skill of the surgeon.

Therapy does not work like that.

In almost every other branch of medicine, a treatable condition presents in patterns a specialist has seen before, and credentials function as a reasonable proxy for outcomes. A well-credentialed cardiologist is likely to produce better outcomes than a less-credentialed one across most patients. The same pattern holds across the rest of medicine. Experience translates. Specialisation translates. The patient's job is to find the right doctor; the doctor's job is to apply the treatment.

Therapy breaks this structure at the root. The treatment isn't a procedure applied to a condition. It is a relationship between two people, inside which a slow and highly personal process unfolds over months or years. Depression in a 28-year-old queer woman in Chennai processing family rejection is not the same condition as depression in a 55-year-old businessman in Ludhiana processing a decade of financial precarity. The diagnostic label on the intake form will be identical. The actual work will not be.

MindCare by Plum

The research on this is unambiguous. Decades of meta-analyses on psychotherapy outcomes, synthesised in Bruce Wampold and Zac Imel's work (Wampold & Imel, 2015) and in Adam Horvath and colleagues' meta-analyses of the therapeutic alliance (Horvath et al., 2011), have shown that the quality of the therapeutic relationship predicts outcomes more reliably than the specific technique the therapist uses. What the therapist does during the session matters less than whether the client and therapist can work together at all.

Which means the match itself is the primary clinical variable. And most of what directories currently sort therapists by (years of practice, degrees held, listed specialities, star ratings) predicts outcomes only weakly. A directory can be fully populated with excellent therapists and still useless to the specific client in front of it, because what it shows is not what matters.

The evidence of this failure is in the retention curve. Roughly four in ten Indians who book a first session with a therapist on a conventional platform never return. The number is stable across the industry. The response has been to publish more credentials, offer more filters, extend the lists, and ask clients to self-select harder. None of it has worked, because the information clients need to match well is not the information directories carry.

The match itself is the primary clinical variable. What directories sort by is not what matters.

So what actually matters? There are five axes of fit, and each one corresponds to a specific design choice inside Plum's MindCare.

Cultural context

A client processing the end of an arranged marriage is doing different work from a client processing a divorce they chose. A client working through guilt about not caring for ageing parents brings a different weight than a client working through similar guilt with no cultural expectation attached. A therapist trained on Western textbooks will miss load-bearing detail. One trained in India but unfamiliar with a specific regional or communal family dynamic will miss it too. Patients consistently prefer ethnoculturally matched therapists (Cabral & Smith, 2011), and when clients perceive their therapist as culturally competent, the effect on outcomes is substantial. A 2018 pair of meta-analyses by Alberto Soto and colleagues found that culturally adapted interventions improve outcomes (d = 0.50 across 99 studies) and that client-rated therapist cultural competence correlates strongly with treatment outcomes (r = 0.38).

How MindCare addresses this: The clinical panel is built for India rather than imported into it. Every therapist on the panel has deep training in Indian family systems, from arranged marriage dynamics to multi-generational household expectations, inter-caste tensions, and the specific weight of community obligation. At intake, the matching engine asks the client about their cultural background and family structure as inputs to the match, not as filter criteria on a profile. The engine then weighs condition-specific client satisfaction data from past clients who presented similarly, to identify therapists whose track record shows capacity in the specific context, not just the paper credentials.

Language

Many Indian clients use more than one language in a day. Research on bilingual therapy by Jean-Marc Dewaele and Beverley Costa (Dewaele & Costa, 2013) has shown that emotional access in a second language is often shallower than in the first: a client can analyse their childhood in English and still only cry about it in Hindi. A session conducted entirely in English can produce articulation without reaching the layer where grief or anger live.

How MindCare addresses this: Language preference is an explicit input to the matching engine, not a filter checkbox on a profile. The panel includes clinicians trained in multiple Indian regional languages, and the engine matches on actual linguistic range — including the capacity to follow code-switching mid-session — rather than a declared language on a form. A client who thinks in English and feels in Hindi is matched to a therapist who can work across both, not to one who speaks both.

Age and life stage

A client in their twenties working through the gap between who they expected to become and who they are now is doing different work from a client in their fifties working through the gap between who they became and who they had hoped to be. A therapist who has been close to the client's life stage brings a recognition that a distant clinician cannot.

How MindCare addresses this: The engine treats demographic fit, including age proximity and life-stage match, as a weighted input. It is never the only variable. A brilliant therapist 20 years older than the client can still be the right match, and the engine reflects that. But life stage is treated as a signal rather than ignored, and clients booking their first session can state life-stage preferences at intake.

Gender

‍A woman processing years of harassment at work has a specific therapy problem that a man cannot easily follow through technique alone. The therapist's own gender identity, lived experience, clinical training on gendered dimensions of the work, and comfort sitting inside those conversations all shape how the conversation plays out.

How MindCare addresses this: Clients state their preference on therapist gender at intake, and the matching engine treats this as a primary signal. More than 70% of the MindCare panel are women. The panel includes clinicians with specialised training in gendered trauma, workplace harassment, reproductive loss, and the specific psychology of women navigating Indian corporate cultures. None of this is visible on a directory profile.

Sexuality

‍A queer client coming out at 30 after a lifetime of masking is a standard case on an Indian workplace mental health platform now, not an edge case. Many Indian therapists were trained in frameworks where queer identity was treated as a problem to be worked through rather than an identity to be affirmed. John Pachankis and colleagues' 2015 randomised controlled trial of LGB-affirmative cognitive-behavioural therapy for gay and bisexual men (Pachankis et al., 2015) found that affirmative treatment reduced depression and anxiety more effectively than standard approaches, and a growing body of work has extended similar findings to other sexual minority populations.

How MindCare addresses this: Clients can state sexuality-related preferences at intake, including requesting an explicitly queer-affirming therapist. The panel includes clinicians with specific training in LGBTQ-affirmative practice, polyamory-aware care, and work with clients processing family or religious rejection after their orientation became known. The engine weighs this specialisation against condition-specific client satisfaction, so a queer client at 30 is matched to a therapist whose historical outcomes with similar clients are strong, not to the first available clinician on the roster.

Each of these five axes is a design choice, not a feature. The matching engine is built on all of them at once.

Matching is the first design problem a platform like MindCare has to solve. It isn't the only one. Four other decisions sit alongside it.

Note continuity, progress tracking and one-tap switching

‍Even strong matching occasionally produces a wrong fit. When it does, a client can switch therapists in one tap, and the clinical notes carry over. The new therapist starts with context. The client is not asked to retell their childhood from the beginning. This design response addresses one of the biggest hidden costs of conventional platforms: the retelling tax that keeps clients attached to the wrong therapist long after they know the match has failed, or makes them abandon therapy entirely instead of starting over.

Progress is tracked with validated clinical tools rather than impressions. MindCare uses the PHQ-9 for depression, the GAD-7 for anxiety, and the SRS (Session Rating Scale) to measure the therapeutic alliance itself after each session. These instruments are the same ones used in randomised clinical trials, which means a client's progress is measured against a scale the research literature can validate rather than against a therapist's subjective sense of how it's going. It also means a stalled case can be identified early. If PHQ-9 scores plateau for a depression client at the four- or five-session mark, the clinician and the matching engine both have the data to know a change is needed.

The integrated physical health check

‍Low mood, exhaustion, brain fog and persistent anxiety are among the most common reasons employees book a first therapy session. They are also among the most common presentations of conditions that have nothing to do with mental health. Thyroid dysfunction can look like depression. Low vitamin D or B12 can look like exhaustion. Iron deficiency anaemia can look like generalised anxiety. For a client sitting opposite a therapist in session one, there is no way to tell which problem they are actually describing.

MindCare is the only workplace mental health platform in India that closes this gap inside the same care pathway. Alongside therapy access, the programme includes a GP and psychiatrist review and a 57-marker blood test designed specifically for mental health presentations. The panel covers thyroid function, vitamin D and B12, iron studies, blood sugar, stress hormones, minerals, inflammation markers and liver function. The tests follow WHO and ICMR guidelines, samples are collected at home, medicines are delivered to the client's door, and a doctor reads every report before it goes back into the care pathway. On a conventional workplace programme, each of these functions (therapy, psychiatry, GP consults, lab tests) sits with a different vendor, and the physical causes of mental health symptoms routinely go unchecked. On MindCare they are one flow.

The commercial case is also sharper than most buyers assume. The programme's own data suggests that undiagnosed physical conditions presenting as mental health symptoms cost employers up to thirty additional sick days per employee per year. A case routed through the physical health check and resolved as a thyroid or deficiency issue returns the employee to full functioning in weeks rather than keeping them in therapy for a condition they do not have.

Confidentiality architected under the DPDP Act

‍Individual client data is never visible to the employer. What HR sees is aggregated utilisation at the population level. Session notes, screening responses, and individual identity remain between the client and the clinician. Confidentiality is not a product policy here; it is the legal architecture under India's Digital Personal Data Protection Act, 2023.

Employer-paid access with real-time triage

‍Employees never pay for a session, which removes the single most common blocker to workplace therapy. And when a client presents in crisis, the engine routes them to immediate clinical attention rather than through the standard matching flow.

Family coverage

‍An employee carrying a partner through postpartum depression, a teenage child through exam anxiety, or an ageing parent through grief cannot do their best work while that person is struggling at home. MindCare extends the same care pathway to the family members an employer chooses to cover, including spouse, children and parents, with no session caps on any of them. Each family member is matched to their own therapist, so a mother and her twenty-year-old daughter are not sharing a clinician. Couples sessions are available when the pressure is on a relationship rather than on one person, and family sessions are available when the whole household needs to be in the room.

This matters commercially for a reason most workplace mental health programmes miss. The leading external cause of an employee's own disengagement is often not their own mental health but a family member's. A parent with dementia, a child with ADHD, a sibling in crisis. Platforms that stop at the employee and treat the household as out of scope keep paying for the employee's absence without ever addressing what is producing it. MindCare's design assumes, correctly, that the household is the unit of care.

The engagement data reflects what all of this produces. On MindCare, 57.1% of members remain engaged through Session 3, against an industry standard of around 15%. Among returning users, 55% rebook within 7 days of their last session and 17% rebook on the same day. 80% of bookings are self-referrals. 90% of users rate their experience positively. The retention curve looks different because the design is different.

The practical lesson for anyone buying workplace mental health is that the vendor matters less than the model the vendor operates on. A platform with a brilliant directory and no matching will underperform a platform with a modest directory and strong matching, because what makes therapy work is not credential richness. It is fit, legally protected, and designed around the specific person rather than an abstract category.

Further reading

‍