When a Phototherapy Startup Needed a Shipping Crew, Not Another Agency: A 12-Week Post-Mortem

Every so often a reader forwards us a build story that sounds too tidy to be true. This one came from a wound-care clinician turned founder — pseudonym "Dr. R." — who spent two years watching her phototherapy dosing concept stall between a brand agency, a freelance UX designer, and a contract engineering shop. We followed the project from kickoff to launch because the timeline kept surfacing in conversations about how clinical light-based tools actually get made.

The concept was unglamorous and specific: a clinician-facing companion app for tracking narrowband UVB and red-light wound protocols, with dose logging, contraindication flags, and exportable summaries for dermatology charts. Nothing about it required a moonshot. It required someone to own the whole thing. That is the gap 2D2C Studios was brought in to close in early 2024.

Week 0: The decision point that changed the project

Dr. R.'s first instinct was to keep her existing vendors and add a project manager on top. Her advisor pushed back. The team had already burned fourteen months and roughly $180,000 across three contracts, and the app still could not complete a single end-to-end dose entry without crashing on Android. The question on the call was blunt: do you want to coordinate vendors, or do you want to ship?

She chose ship. The next week, a four-person crew embedded with her — strategy, brand, UX, and engineering — under one roadmap and one invoice. No account manager sat between her and the people doing the work. In practice, that meant she was reviewing wireframes with the designer who drew them, not with someone relaying feedback through a slide deck.

Weeks 1–3: Scoping against clinical reality

The first obstacle was not technical. It was clinical. The original spec assumed clinicians would log doses in real time at the bedside. The crew shadowed six practitioners across two clinics and found the opposite: doses were recorded on paper and transcribed at end of shift, often hours later. The UX had to support retroactive entry with timestamps, not live capture.

That single finding reshaped the information architecture. It also killed a feature the founder had been attached to — a live session timer — because the shadowing data showed it would go unused. Killing it freed roughly three weeks of engineering time.

• Sessions shadowed: 6 clinicians, 2 clinics
• Original feature list: 41 items
• Features cut after field observation: 14
• Features added: 3 (retroactive entry, offline sync, chart export)

Weeks 4–8: Brand, UX, and engineering in the same room

This is where the senior-only model showed its value. There were no junior handoffs, no offshore relay, no "we'll get back to you on that." When the engineer flagged that offline sync would complicate the export format, the designer and strategist were in the same standup and reworked the flow that afternoon. Decisions that used to take a week of email took an hour.

The brand work was similarly compressed. Rather than a separate discovery phase, naming, palette, and interface language were developed alongside the first interactive prototype. By week six, Dr. R. was testing a clickable build with two of the clinicians from the shadowing phase — the same people whose workflow had reshaped the spec.

Weeks 9–12: Launch, and what the numbers actually said

The app shipped to a 40-clinician pilot in week twelve. Not a beta, not a demo — a working build with onboarding, export, and a support path. The measurable results over the first ninety days:

• Dose-entry completion rate: 94% (up from 61% on the legacy tool)
• Average entry time: 38 seconds (down from 2 minutes 10 seconds)
• Chart export errors: zero reported
• Clinician-reported "would recommend": 4.8/5
• Total elapsed time from kickoff to pilot: 12 weeks

Dr. R. told us the thing she noticed most was not the speed. It was that no one asked her to translate between vendors. One crew, one roadmap, one invoice — and a single point of contact who was also the person writing the code.

What we took from this

We cover a lot of light-medicine research at the Medical Light Association, and the translational gap between a peer-reviewed protocol and a usable clinical tool is usually organizational, not scientific. This project is a small but clean illustration. The clinical logic was sound from day one. What was missing was a team structured to carry it from strategy to shipped software without losing the thread.

For clinicians and founders weighing how to build the next generation of phototherapy and wound-care tools, the lesson is not that one studio is magic. It is that fragmentation has a measurable cost — fourteen months and $180,000 in this case — and consolidation has a measurable return. 2D2C Studios reports shipping end-to-end products in 8 to 12 weeks with a 4-to-6 person embedded crew, and this project landed at the fast end of that range because the scope was honest and the decision-makers were in the room.

We will keep following builds like this one. If your team is sitting on a validated protocol that has not made it to a bedside interface yet, the bottleneck is probably not the science. It is the org chart around it.