The most expensive part of private practice is rarely the part you trained for. It's the writing that happens after the session ends — the reports and notes that pile up on evenings and weekends. For many clinicians a single assessment report runs 10 to 15 hours, almost none of it billable. It isn't slow writing. It's a broken process.
I'm Ian, and I run Soma Health. Over the past couple of years I've spent more time than I can count listening to psychologists, therapists, and counsellors talk about their week. Almost every conversation lands in the same place — the work is good, the clients are worth it, but the documentation is quietly eating them alive. This piece is my attempt to lay out the whole problem honestly, and to say plainly what actually helps.
The after-hours writing few people ever see.
Why is report writing such a quiet problem in private practice?
It's quiet because nobody sees it. Report writing happens after the client has left, after the office has emptied, often after dinner. There's no audience for it and no applause — just a clinician at a laptop, alone, turning testing and observation into something defensible and clear.
It's also quiet because clinicians rarely complain about it out loud. They chose this work to help people, and writing up a careful report is helping. So the cost gets absorbed silently, week after week, until it shows up as something else — fatigue, a backlog, a Sunday that wasn't really a day off.
When I ask clinicians what the hardest part of their job is, documentation is the near-universal answer. Not the clinical work. Not the hard conversations. The writing that comes after. It's the thing they'd most like to hand off and the thing they trust least to anyone else — because a report has to read like theirs.
The research backs up what I hear. One survey of health-service psychologists found a median of around five hours a week — closer to seven on average — spent on non-billable clinical work, and over half of those surveyed named that administrative load as a significant source of work-related stress. That's not a rounding error. That's a part-time job hiding inside a full-time practice.
How long does a single report really take?
Longer than anyone bills for, and longer than most clinicians admit even to themselves. One psychologist told me a single assessment report regularly takes 10 to 15 hours. Another put it as a ratio that's stuck with me ever since — roughly 14 hours of writing for every 5 hours of testing.
Sit with that ratio for a second. The face-to-face part, the part the client experiences and the part that gets billed, is the small slice. The writing — the synthesis, the structure, the careful wording — is nearly three times larger and almost entirely invisible on an invoice.
And reports aren't the whole story. As one clinician put it to me, "a big part of the unpaid work is the prep before and after a session." It's the note you write at the end of the day, the summary you pull together before the next one, the file you reread so you walk in prepared. None of it is billable. All of it is necessary. It adds up across a full caseload into hours that never appear on a schedule.
The literature on clinician exhaustion points the same direction — high client volumes and long hours wear people down, and a lack of adequate resources and support makes it worse. The hours themselves are part of it, but so is doing those hours without the right tools. I'll come back to that, because it's the part I think we can actually change. If you want to go deeper on the timing specifically, that's its own question worth its own piece — why reports take so long is rarely the reason people assume.
The mechanical assembly is where the hours disappear.
Is it slow writing, or is it the process?
This is the question I most want clinicians to hear, because the answer changes everything. It isn't slow writing. The people telling me a report takes 15 hours are excellent writers — careful, precise, fast when the task is clear. The time doesn't go to a lack of skill. It goes to a process that was never designed.
Think about what a report actually demands. You're holding scores from your instruments in one place, session observations in another, the client's history in a third, and your own clinical reasoning in your head. The real work is assembling all of that into a coherent document — pulling numbers into the right sections, restating the same finding in three places, formatting tables, making sure the summary matches the body. That's mechanical assembly, and mechanical assembly is where the hours disappear.
It gets worse because the tools are scattered. Many clinicians cobble together separate pieces — one app for testing, a template in a word processor, a spreadsheet for scores, sticky notes for the rest — and wish, out loud and often, that it were all in one place. Every handoff between tools is a place to lose time and a place to make a mistake.
So no, it's not a slow-writer problem. It's a process problem. And process problems have process solutions — which is a far more hopeful thing than telling a tired clinician to somehow type faster. The fix isn't working harder on the writing. It's taking the assembly off the writer's plate so the writing — the part that needs a human — is all that's left.
Can a report still sound like the clinician who wrote it?
This is the fear, and it's the right fear to have. The moment you mention helping with reports, every good clinician asks the same thing — will it still sound like me? And they should ask, because a report that sounds generic is worse than no help at all. It's a report you have to rewrite, which means you've paid twice.
Here's what clinicians keep telling me they want: their own voice and their own structure, not a generic template. The way they phrase a finding. The order they move through sections. The particular care they take in a summary. That's not a preference — it's their clinical signature, and it's part of how their referral sources and their clients trust the work.
So the only version of this that's worth building is one where the report still reads like yours. The goal isn't a document a machine would write. It's the document you would write, with the mechanical assembly already done so you're editing and refining instead of starting from a blank page at 9pm. You stay the author. You keep your judgment on every line.
I won't pretend the wording lands perfectly on the first pass every time — no honest founder would. But the standard I hold us to is simple: a report that reads like yours, that you'd be comfortable signing without apology. If the question of whether an AI-assisted report can still sound like you is the one keeping you up, it deserves a full and honest answer of its own, and it's one of the things I most want clinicians to test for themselves rather than take my word on.
Does faster reporting mean lower quality?
The instinct is reasonable — faster usually does mean sloppier. But that's only true when "faster" comes from cutting corners on the thinking. The kind of speed worth wanting comes from somewhere else entirely. It comes from removing the parts that were never the thinking in the first place.
Formatting a table is not clinical judgment. Copying a score from one section into another is not clinical judgment. Restating the same finding in the summary that you already wrote in the body is not clinical judgment. When that mechanical layer comes off your plate, you're not doing less careful work — you're spending all of your care on the part that actually needs it.
I'd argue the scattered, manual process is the bigger risk to quality, not the smaller one. Every late night, every copy-paste between tools, every report written while exhausted is a chance for a real error to slip in. A clinician working at 11pm on their fourth report of the week is not at their sharpest, and they'd be the first to tell you so. Reducing that load protects quality — it doesn't threaten it.
The honest framing is this: speed isn't the goal. Getting your evenings back without compromising the work is the goal. Speed is just what it looks like from the outside when a broken process finally gets fixed.
One place for the notes and the reports, not five scattered tools.
What should clinicians look for in report software?
Look for something that keeps you the author. The single most important test is whether the finished report reads like yours — your structure, your voice, your judgment intact. If a tool produces a generic document you have to substantially rewrite, it hasn't saved you anything; it's just moved the work around.
Look for one place instead of five. So much of the lost time lives in the handoffs between scattered tools, so a real solution puts the notes side and the reports side together — your scores, your observations, your draft — without making you stitch them by hand. The clinicians who tell me they wish it were all in one place are describing the actual feature that matters.
Look for software that recognizes your instruments and your way of working, not one that forces you onto a fixed template. And look for a tool that treats privacy as the starting point, not an afterthought — anything you share to set things up should be de-identified, because that's simply how clinical work should be handled.
Most of all, look for something honest about what it is. Soma is software for clinicians — it helps you draft your reports and notes faster while you keep your clinical judgment on every line. It doesn't practice, it doesn't advise, it doesn't replace you. It takes the assembly off your plate so the judgment is all that's left for you to do. If you want to see exactly how that works in practice, you can take a closer look at the product and decide for yourself.
And if you're weighing the smaller pieces of this — the after-hours documentation that fills your evenings, whether a template can ever really fit your work, why a single report eats a whole day — those are all branches of the same root problem. This is the trunk. The rest grow out of it.
The quiet problem deserves a real answer
If there's one thing I'd want a clinician to take from this, it's that the report-writing problem is real, it's expensive, and it is not a personal failing. You are not slow. The process was never built for you, and you've been quietly absorbing the cost of that for years.
What helps isn't pressure to do more. It's a process that hands you back the hours that were never really clinical to begin with — the formatting, the copying, the assembly — while leaving your voice and your judgment completely yours. That's the whole idea behind what we're building, and it's the standard I want to be held to.
Thank you for the work you do, and thank you for reading this far — it means a great deal to me that clinicians keep telling me the truth about their week. It's the only reason any of this gets better. If you ever want to talk through how report writing actually plays out in your practice, I'd love to hear about it. Reach out and I'll take care of the rest.
— Ian
