Getting the most from a remote studio

A founder once told me, over coffee in Surry Hills, that she'd never hire a remote studio again. Her previous agency had designers in three time zones, and every question took two days to answer. "I'd ask for a small change on Monday," she said, "and by Thursday I'd get a Figma comment asking what I meant."

I don't blame her. But I want to be precise about what actually failed there, because it wasn't geography. It was an engagement designed as if everyone shared an office, run across eleven hours of time difference. Collocated habits — tap someone on the shoulder, whiteboard it, walk over — don't survive distribution. They have to be replaced, deliberately, with rituals that are often better than the office versions.

Brassfern has been distributed since our second year: Sydney, Melbourne, Wellington, Auckland, Singapore, London. Here's the buyer's-eye view of how to get the most from a studio like ours — and how to tell, before you sign, whether a remote agency has actually solved this or is hoping you won't notice.

Ask how they design the overlap

The first thing to ask any distributed studio is not "where is everyone?" but "how do you use the overlap?" Two to three hours of shared working time is plenty — if it's spent on the things that genuinely need synchrony: decisions, disagreements, and demos. It's a disaster if it's spent on status updates.

On our engagements, the overlap window is protected for three rituals:

A weekly demo, never a status call. Fridays at Brassfern are demo day — the client sees working software or real screens, every week, without exception. Status is what the demo is. If nothing can be shown, that is itself information, and it surfaces a full week earlier than a RAG report would.
One decision meeting with a pre-read. Anything that needs a client decision arrives in writing twenty-four hours ahead: the question, the options, our recommendation, the cost of being wrong. The meeting then argues about substance instead of absorbing context. Decisions that used to take three meetings take one.
Open office hours. A standing slot where anyone on the client side can drop in with the "small question" that would otherwise marinate for two days. Most drop-ins last four minutes. That's the tap on the shoulder, rebuilt.

When you're evaluating agencies, ask to see a recent decision pre-read (anonymised is fine). If they don't write them, their remote practice is improvisational.

Demand written defaults — and read how they write

Distributed teams live or die on writing. Not beautiful writing — decisive writing. The test is whether a question left asynchronously comes back with a recommendation, or with more questions.

Our rule, which you should hold any studio to: every async message of consequence ends with a default. "We recommend option B, shipping Wednesday unless you object by Tuesday 5pm." This one habit is worth more than any tool. It converts time zones from a tax into a feature — clients wake up to progress with a decision teed up, not a thread of "thoughts?"

Red flags to watch for during the sales process:

Proposals that take a week to answer a pricing question (their internal async is broken).
Long video calls where the deliverable could have been a page (they mistake meetings for work).
Beautiful decks, thin written follow-ups (they present better than they operate — and you'll live in the operations).

Design the decision turnaround on your side too

Here's an uncomfortable truth we tell every prospective client: most remote engagements slow down because of the client's turnaround, not the agency's. A distributed studio will front-load questions in writing; if those questions sit unanswered for four days, the critical path just grew by four days, and no amount of timezone cleverness recovers it.

So we agree on turnaround norms up front, and we put them in the statement of work:

Standard decisions: two business days, else the studio's recommendation ships.
Blocking decisions: flagged explicitly as blocking, next business day.
One decision-maker by name. Not a committee, not "the leadership team" — a person, with a named deputy. Poetry dies in committees; software just dies.

This isn't the agency being precious. It's the single biggest lever on project pace, and it's entirely on the client side. The engagements where a founder answers our Tuesday pre-read by Wednesday lunch are the ones that ship early. Our pricing page is honest about this: fixed-scope pricing assumes this cadence, because scope certainty and decision latency are the same number wearing different hats.

Get the tooling boring and the security explicit

Tool sprawl is where remote projects leak. Our default stack is deliberately dull: one shared channel, one task board, one design file, one doc home. Every artefact has exactly one place it lives, and the one-pager says where. If an agency pitches you a bespoke Notion-and-Linear-and-Basecamp-and-also-email arrangement, they're describing their chaos, not your system.

On security, ask the unglamorous questions early:

How are credentials shared? (Answer should involve a password manager with sharing, never email or chat.)
Who has access to your analytics, your CMS, your repository — named individuals, auditable, revocable at end of engagement.
What's the device policy for contractors touching your data?

A studio with crisp answers has been asked before. A studio that says "we're very security-conscious" and reaches for a coffee has not. Our handover playbook covers how this unwinds cleanly at the end — the end is where you find out how the access was run all along.

Visit once, even if nobody has to

This sounds like a contradiction in a piece about remote work, so let me be exact: remote-first doesn't mean never-together. We budget a kickoff in person wherever logistics allow — two days, at the client's office or ours — because the first forty-eight hours of a relationship do work that no document does. You learn how people argue, what they care about, which jokes land.

After that, remote is the default and nobody flies for a status meeting. The ratio matters: roughly 5% together, 95% distributed is a dramatically better allocation than either extreme. Fully-remote-from-day-one engagements work, but they take a month longer to develop the shorthand. Fully-onsite engagements are collocated theatre that evaporates the moment we leave the building.

What to put in the statement of work

If you're buying, here's the remote-readiness clause list we'd happily sign (and wrote):

Named squad with time zones listed. You should know exactly who is doing the work and when they're awake. Our team page names the humans on the tin; your SOW should too.
Weekly demo commitment with a definition of "demoable" (working software or real screens — not a report about software).
Decision turnaround norms, both directions — the agency's response times and yours.
The async default rule in writing: recommendations with deadlines, not open questions.
Overlap window protection — the two to three sacred hours, agreed in both calendars.

Any studio that flinches at these terms is asking you to fund their learning curve.

The honest trade-offs

Remote done well buys you senior people who'd never relocate to your city, deeper written thinking, and a pace advantage once the machine is humming. It costs you spontaneous osmosis — the overheard conversation that solves a problem — and it punishes ambiguity harder than an office does. Vague briefs survive in hallways; they die in writing. Which is, in the end, a feature too.

The founder from the coffee shop? She signed with us a month later. Her feedback at the end of the engagement was the line I'd put on our studio page if it weren't immodest: "It felt less remote than the agency two suburbs away."

That outcome isn't magic. It's rituals, defaults, and a refusal to let status meetings substitute for work.

Key takeaways

Geography doesn't break engagements; imported office habits do. Judge a remote studio on its replacement rituals, not its time zones.
Protect the overlap for decisions, disagreements and demos — never for status.
Demand written defaults: every async question returns with a recommendation and a deadline.
The biggest pace lever is usually the client's own decision turnaround. Agree norms in the SOW, with one named decision-maker.
Keep tooling boring, security explicit, and budget roughly 5% of the engagement as in-person time, front-loaded.

Frequently asked questions

How many hours of time zone overlap do we actually need? Two to three, spent well. More overlap is nice to have; less is workable if the written culture is strong — we've shipped with clients nine hours apart. What you can't survive is overlap spent on status theatre while decisions die asynchronously.

Should we insist on daily stand-ups with the agency? No. Stand-ups are an internal coordination tool for the squad; making clients watch them is transparency cosplay. What you need is the weekly demo and a shared board you can read any time. If you want more signal than that, the answer is a better board, not more meetings.

What if our team is fully in-office and prefers calls? Then a good remote studio will meet you part-way — but expect us to convert the important calls into written decisions afterwards. That's not bureaucracy; it's the record that keeps a distributed build honest. If your culture refuses to write anything down, a remote engagement will feel harder than it should, and a local studio genuinely might serve you better. We'd rather tell you that than discover it in week six.

How do we build trust with people we've never met? Fast, small, verifiable commitments. Agencies earn remote trust the same way software does: by shipping what they said, when they said. In the first fortnight, watch the micro-promises — "you'll have the pre-read by Thursday" — and check them. A studio that keeps tiny promises keeps big ones. The inverse is also diagnostic.

Can a remote studio handle a crisis at 2am our time? With the right rota, yes — and this is worth asking about explicitly for anything revenue-critical. Our support arrangements define severity, paging and who answers, in contract language, because "we'll jump on it" is not a rota.