Async updates & visibility

The default failure is the stream-of-consciousness update: a paragraph of what you did, ending with something ambiguous. The reader has to extract the state, guess whether to worry, and ask a question — which costs a full timezone round trip. Invert it: conclusion, then evidence, then the ask.

A weekly or daily update that works has four lines: status (on track / at risk / blocked, chosen honestly), what changed since last time, what's next, and what I need from you, with names attached. "At risk" written early is a gift to a manager; "at risk" written the day of the deadline is a problem handed over. Raising a risk before it's certain is a senior behaviour, not an admission — and the way to raise it is with the mitigation attached: "the vendor API is slower than documented; I'm timeboxing a workaround until Thursday, and if it doesn't land we cut the export feature from v1."

Escalation follows the same shape. Say what's blocked, what you've already tried, what you're asking for specifically, and by when. A blocked engineer who waits quietly for two days has made a decision on the team's behalf.

For questions, batch them and give context: what you're trying to do, what you've tried, what you think the answer is. "Is it A or B?" with your own recommendation gets an answer in one message; "how does auth work?" gets a meeting.

Write in the open — a channel or a doc rather than DMs — because the same answer then serves everyone else who'd have asked, and it leaves a searchable record. Add the decision back to the ticket or doc when it's made in a call; a decision that lives only in someone's memory of a Zoom call effectively doesn't exist.

And know when async is the wrong tool. Anything with emotional weight, a real disagreement, or more than two rounds of confusion is faster and kinder on a call. The rule: decide synchronously when it's contentious, record asynchronously always.

At remote-first companies — which is most of the target list — written communication is the job as much as the code is, and interviewers read your take-home README and your PR descriptions as samples of it. Managers explicitly cite "I always know where their work stands" as a senior differentiator.

• Lead with the conclusion; supporting detail goes after, for whoever wants it.
• A status update states status, change, next, and the specific ask with a name attached.
• Flag risk early with a proposed mitigation — early bad news is useful, late bad news is a problem.
• When blocked, say what you tried and what you need by when; silent blockage is a decision.
• Ask questions with context and your own proposed answer to collapse the round trips.
• Communicate in public channels and docs so answers are reusable and searchable.
• Write decisions back into the ticket or doc after any synchronous conversation.
• Switch to a call for disagreement, emotion, or the third round of confusion.