Based in Kyiv

©2026

Three Weeks of Async Feedback With a Studio We’d Never Met in Person

An open laptop on a desk at night with city lights behind

Last spring we built a site for a branding studio in Warsaw without a single call before week four. Here is what that actually looked like, hour by hour.

The brief came through a mutual contact in late February. Marta runs a three-person branding studio called Prosto out of a converted warehouse in Praga, Warsaw, and they needed a site for a furniture client’s launch in six weeks. We never scheduled a kickoff call. Instead she sent a Notion doc with forty bullet points, a Dropbox folder of reference images, and a line that said “let’s just start and talk when there’s something to react to.”

Why async became the default

Part of it was practical. Marta’s studio works mostly with clients in Germany and the UK, so their days are already fragmented across time zones. Adding a call with us in Kyiv, an hour ahead of them most of the year, felt like one more meeting nobody had energy for. Part of it was also a bet: that writing things down forces clearer thinking than talking them through out loud.

We were skeptical at first. We like the friction of a live conversation, the way a half-formed idea gets corrected in real time. But three weeks in, we had a folder of forty-one Loom videos and a revision history that told a more honest story than any meeting notes would have.

A phone face down beside a mug and a folded note in morning light
No call before week four, so every question waited for the morning.

The actual rhythm of the week

Monday: the drop

Every Monday morning we pushed whatever we’d built over the weekend to a staging URL and recorded a six to nine minute Loom walking through decisions. Not just “here’s the homepage” but why the hero used a 4:5 image crop instead of 16:9, why the product grid broke to two columns at 760px instead of the more common 768px.

Wednesday: the reaction

Marta and her designer Tiago would watch the video, leave comments directly in Notion against timestamps, and sometimes record their own two-minute Loom back. The best feedback we got on the whole project was a 47-second clip of Tiago scrolling in silence, pausing twice, saying nothing. We knew exactly which two sections weren’t working.

Friday: the close

We’d wrap the week with a changelog, a flat list of what shipped, what’s pending, and one open question for the following Monday. No more than five lines. Keeping it short meant it actually got read.

What broke down

Not everything worked. Tone is hard to read in a bulleted comment. Twice we misjudged a note as a hard requirement when it was really Marta thinking out loud, and built something she then had to ask us to undo. We eventually agreed on a small vocabulary to flag intent: “FYI” for observations, “ASK” for real requests, “BLOCKER” for anything holding up the next step.

Design decisions that needed real debate also suffered. There’s a spacing argument in the product detail page that took eleven back-and-forth comments to resolve, something a five-minute call would have settled in two.

The async model is fast for execution and slow for disagreement. If you can’t disagree quickly, you ship the wrong thing slowly instead.

What we’d keep doing

  • Recording a walkthrough video every time, even for small updates, because it forces us to justify decisions out loud
  • Writing a five-line changelog at the end of each week instead of a long status report nobody reads
  • Agreeing on a shared vocabulary for feedback intent before the first real disagreement happens
  • Keeping one weekly sync call as a pressure valve, even if it’s only fifteen minutes

The number that surprised us

By the end of the project we counted the hours differently than we expected. Across six weeks we spent roughly 14 hours in actual meetings, including the one real video call we eventually had in week four to sort out the navigation structure. Compare that to a typical six-week site build with weekly one-hour calls, which lands closer to 20 to 24 hours of live meeting time. The difference wasn’t just saved time. It was saved attention, because every minute of async communication was something somebody had chosen to write down.

Would we do it again

Yes, but not for every project. It worked here because Marta’s studio already thinks in documents, and because the scope was tight enough that ambiguity had somewhere to land. For a messier brief, or a client who needs to see conviction in someone’s face, we’d push for calls earlier. The site shipped on time, the furniture client was happy, and we still haven’t met Tiago, only his eleven seconds of silent scrolling, which taught us more than most calls do.

Leave a Reply

Your email address will not be published. Required fields are marked *

Recent posts