Ask therapists what they want in an EMR and the answer is consistent: a fast, clean place to write the clinical note, with billing and coding running quietly in the background instead of a database they have to fight. They want far fewer clicks, a record that is genuinely useful for clinical thinking, and AI that handles the admin rather than acting as the therapist. The pain behind that wish list is measurable. Documentation is the top driver of therapist burnout, clinicians spend roughly two hours in the record for every hour of direct patient care (Annals of Internal Medicine), and after-hours charting is tied to significantly higher burnout (Academic Medicine, 2025). This is the full pain list, and the EMR therapists describe when you actually ask them.
I am a licensed clinical social worker, and I build software for clinicians, so I get this in both directions — the venting and the wishful "if I could just have." This is what I hear, organized into the two questions that matter: what is the pain, and what would actually fix it.
The pain, named honestly
It was built for billing and checkboxes, not for therapy. This is the root of all of it. The dominant systems are optimized for billing, compliance, insurance audits, and data extraction — not for clinical work. Everything gets forced into structured fields, templates, and redundant attestations, and what should be a clean clinical note becomes a bloated compliance artifact. The note stops being a record a clinician writes and becomes a form a clinician feeds.
The "too many clicks" tax. Simple, routine tasks require an absurd number of clicks, pop-ups, mandatory fields, and navigation steps. The most common refrain — you hear it everywhere clinicians talk to each other — is that these tools were obviously designed by people who have never done the job. That is not a UI nitpick. It is a per-encounter tax on attention that compounds across a full caseload.
Note bloat, and the death of the useful record. Notes get long, copy-pasted, and depersonalized until the official note is no longer the real clinical record. So clinicians keep the real one somewhere else: a Word document, personal notes, a spreadsheet. When a sizable share of your providers are maintaining a parallel record because the EMR is too rigid to think in, the EMR has failed at its one job — being the place the clinical truth lives.
Behavioral-health-specific friction. Therapy carries documentation realities that hospital-style platforms handle badly. Psychotherapy process notes get extra protection under HIPAA and are meant to be kept separate from the rest of the record;[1] progress notes are a different artifact with different rules. Layer on session-timing codes, treatment-plan cadence, and insurance requirements, and documentation starts to feel punitive on top of work that is already emotionally heavy.
The time sink, and the burnout that follows. The end state is the 9pm note. Providers routinely describe spending more time feeding the EMR than being present with clients, and the research backs the lived experience. A landmark time-and-motion study in the Annals of Internal Medicine found clinicians spend roughly two hours on the record and desk work for every hour of direct patient care.[2] A 2025 study of nearly 9,700 resident physicians found that high after-hours "pajama time" in the record was associated with significantly higher odds of burnout.[3] In behavioral-health practice surveys, documentation is named the single biggest driver of burnout, ahead of caseload and pay. The charting-after-dark problem is measured, not anecdotal.
What they actually want, when you ask
Here is the encouraging part. The wish list is not a fantasy of more features. It is a plea for less — and it is strikingly consistent from one clinician to the next.
A word processor that happens to handle billing. This is the whole thing in one line. Clinicians want clean, fast, narrative writing with minimal friction, and they want the compliance and coding machinery to run quietly in the background instead of in their face. The note is for thinking. The billing is plumbing. Today those priorities are inverted; clinicians want them put back in order.
A dramatically lower click count. Fewer mandatory fields, fewer pop-ups, fewer navigation steps for the things done a dozen times a day. Every click removed from a routine task is attention returned to the person in the room.
A clean, modern, intuitive interface. Something that feels like current software, not enterprise database software from 2007. This sounds cosmetic and is not: an interface a clinician can move through without friction is the difference between a tool that disappears into the work and one that sits on top of it demanding to be managed.
A record that is actually useful for clinical thinking. Easy to find past notes. Easy to see a client's patterns over time. Easy to write in a way that supports real therapy work rather than insurance defense. The EMR should make a clinician a better clinician by surfacing the right history at the right moment — not bury the signal under copy-pasted boilerplate.
Respect for the therapy workflow. The distinction between process notes and progress notes, handled correctly and protected correctly. Session timing that fits how therapy actually runs. A treatment-plan cadence that is a clinical rhythm, not an arbitrary form due date. The relational nature of the work, reflected in the structure of the tool instead of fought by it.
And underneath all of it: clinicians want something lightweight and clinician-first, not the heavy hospital-style platform that got forced onto an outpatient practice it was never designed for.
Where AI actually fits in this picture
Notice what is not on that wish list: an AI that does therapy. The interest in AI among practicing clinicians is specific and almost entirely about the administrative side — AI that can handle the documentation and the billing so the human can stay in the clinical note. That is the honest, evidence-supported use, and it is the opposite of the "AI will see the patient" pitch. I have written separately about what therapists actually want from AI, and why it isn't a chatbot therapist; the same instinct shows up here. Let the machine eat the paperwork. Keep the clinician in the room and on the note.
Done right, that looks like a note drafted from a consented session record — which the clinician edits, corrects, and signs — with coding and claims handled in the background and surfaced only when a human decision is actually required. Studies of ambient documentation tools have reported real reductions in documentation time, after-hours charting, and self-reported burnout when implemented well.[4] The clinician stays the author and signer of every clinical artifact. The AI is the staff, not the provider.
The EMR therapists describe
Put the pain and the wish list together and the specification writes itself. The EMR therapists actually want is a clean, fast place to write the clinical note, where the note is the point and the billing runs quietly underneath it. It has a low click count and a modern interface. It treats the record as a thinking tool — searchable, pattern-aware, genuinely useful at the start of the next session. It respects the realities of therapy: process notes versus progress notes, session timing, the protected and relational nature of the work. It uses AI to absorb the administrative load, never to impersonate the clinician. It is lightweight and clinician-first. And it does not hold a practice's data hostage when the clinician decides to leave.
That EMR mostly does not exist yet, which is exactly why so many capable clinicians have quietly given up and gone back to their Word documents. It is also the thing I am building toward — not because it is a clever product idea, but because it is the tool I, as a working clinician, would actually want to use on my own caseload. The note should write itself from the session. The clinician should still be the one who reads it, fixes it, and signs it. The billing should run in the background and tell the truth. And the hour after the last client logs off should go back to belonging to the clinician.
If you are a therapist who has made that quiet peace with your EMR — the shadow document, the 9pm notes, the resignation — you are not asking for too much. You are describing software that should already exist.
So here is the call. The EMR that finally treats the note as a thinking tool and the billing as plumbing is going to take the independent-practice market. Not the biggest platform. The one built for the clinician. We are building that one, and we intend to be the reason you finally close the Word doc for good.