TL;DR.
A self-directed audit of an existing therapy app — booking, appointment management, specialist listing. The redesign focused on removing friction where it hurt users and adding the right friction where actions had consequences.
Why this app, why an audit.
Take Therapy is a digital therapy app — appointments, specialists, session management. I'd been spending time inside it as a user, and a few moments kept snagging me: a cancel button sitting one tap away from confirmed sessions; specialist cards that asked me to drill in just to find out a price; a booking form that gave no sense of how long it would take.
So I took the time to audit it properly — not from inside the company, just as someone who cares about how this kind of product feels. The output is a redesign proposal, not a shipped change. Every decision below traces back to either a user observation I made firsthand or a UX principle that explains why the existing pattern was costing trust.
Four specific friction points.
These aren't subjective gripes — they're places where the interface fights against either user safety or user understanding. I documented each one against the screen it lived on so the redesigns later could be specific, not aspirational.
A cancel action for an upcoming session sat one tap away on the home screen, with no confirmation modal. A misfire could vacate a booked specialist slot.
Cancel and reschedule rendered with the same visual weight as primary actions — same colour family, same size, no semantic warning.
Specialist listing cards showed name + photo + speciality, but pricing and ratings sat one tap deeper. Two of the three things that drive a booking decision required an extra step.
The booking form collected everything on one screen — concern, specialist, time, payment, notes — with no progress indicator. Users had no read on how close to done they were.
Four principles, four decisions.
Every redesign in this audit is justifiable in two sentences: what problem it solves, and which named principle says the existing pattern was wrong. I leaned hardest on these four:
Decision time scales with the number of options visible at once. The single-page booking form violated this — every choice present, no chunking.
Hit accuracy scales with target distance + size. The proximity of a destructive cancel button to common actions on the home screen was a Fitts'-law-shaped foot-gun.
Productivity drops if a system takes more than ~400ms to respond. Decision-critical info (price, rating) being one tap deeper than it needs to be is a slow-feedback problem in disguise.
Nielsen's 5th heuristic: prevent errors at the source rather than report them after. Destructive actions without confirmation are the cleanest violation of this heuristic.
The redesign, screen by screen.
I didn't redesign the whole app. The audit was scoped to the four flows where the friction was costing users something real — home, specialist listing, booking, and appointment management. The proposals below are deliberate, not exhaustive.
Both surfaced on the card itself. No drill-in required to know whether a specialist is in budget or well-rated. Doherty.
Treated as a filterable attribute, not body copy — supports scanning across many specialists fast.
A small 'next slot' hint replaces having to tap in just to see the calendar. Reduces dead-end drills.
Three labelled steps replace the single-page form. Visual contract for 'how long is this going to take.' Hick.
Concern → specialist + time → confirm. Each screen owns one decision, not five. Hick.
A bottom sheet on each step shows what the user has chosen so far — no need to backtrack to verify.
The cancellation flow change is the clearest UX-law win in the audit: the combination of Fitts (more distance to the destructive action) and Nielsen-5 (explicit confirmation) means you have to genuinely intend to cancel for cancellation to happen. The cost is one extra tap. The trade is right.
Honest scope of what this audit doesn't prove.
This is a redesign proposal, not a shipped change with measured impact. I'm not going to invent numbers. What this audit does claim is that each change has a principled reason — not "it looks better," but "the existing pattern violates a specific named law and here's the version that doesn't."
The next honest step is putting the redesigned flows in front of real users in a moderated test and seeing whether the booking-form chunking actually reduces drop-off, whether the destructive-style confirm materially changes accidental cancels, and whether surfacing price on the listing card improves time-to-first-booking. Those are testable. Without running that test, the claim stops at principled — which is the honest level it should stop at.
- Grounding every change in a named principle made trade-offs explicit. 'This adds friction' is easy to argue against; 'this adds friction because Nielsen-5' is not.
- Scoping to four flows kept the audit shippable as a proposal — a 40-screen redesign would have been ignored.
- Walking through the existing app as a user before opening Figma stopped me from solving problems that weren't actually painful.
- I'd pair the audit with one or two real moderated tests on the existing app before redesigning — even three users would have anchored the friction list better.
- The progress-bar choice on the booking flow was made without prototype-testing. I'd want to see whether users prefer a bar, dots, or labelled steps.
- I didn't look at the reschedule flow in this pass, only cancel. Reschedule has the same destructive-action shape and would benefit from the same treatment.
If you're sitting on a product that needs the same treatment — an honest audit of where users are getting stuck, grounded in named UX laws and shipped as a prioritized roadmap rather than a 40-page deck — this is the work I do as UX research and redesigns.
On this case study.
Solo UX audit and redesign proposal. I worked through the existing therapy app as a user, documented friction in the booking, cancellation, specialist-listing, and appointment-management flows, then proposed targeted redesigns scoped to those four flows — not a top-to-bottom rebuild. No external collaborator on the design side; the output was a proposal deck, not a shipped change.
A proposal. Take Therapy is a real app I'd been spending time inside as a user; the audit was self-directed, not commissioned. I'm explicit about that in the case study because I won't claim measured impact I can't prove. The honest claim the audit makes is that every change has a principled reason — a specific named UX law the existing pattern violates — not that it's been A/B tested against the original.
Four named laws, one per friction point. Hick's Law for chunking the single-page booking form into three labelled steps. Fitts' Law for moving the destructive cancel action away from common taps. The Doherty Threshold for surfacing price and rating on the specialist card so users don't have to drill in to find decision-critical information. Nielsen's 5th heuristic (error prevention) for adding a confirm sheet with explicit destructive styling on cancel.
Figma — start to finish. The audit lived as annotated screens of the existing app paired with redesigned screens, sitting side-by-side in the same Figma file so each change was visible against what it replaced. No separate prototyping tool; the proposal is a static deck because the next honest step is moderated user testing on the redesigned flows, not a higher-fidelity prototype.