How do you track project status across a distributed team?

What was actually lost

Ask a leader who managed teams in an office how they knew a project was in trouble, and the answer is rarely "the status report." Several people in our research described the real mechanism:

"Hallway and water cooler conversations. Program managers were my ears and eyes." — group project manager, 100–150 engineers

"Walking the halls." — VP of engineering, 120 engineers

It's worth being precise about why that worked, because the properties are what any replacement has to supply.

It was continuous. Not a weekly sample — dozens of small observations a day, accumulating into a background sense of how things were.

It was free. No preparation, no calendar entry, no cost to either party. Which meant it happened often enough to catch changes early.

It was off the record. "I'm not sure this is going to work" said walking to the kitchen commits nobody to anything. The same sentence in a status field is a statement someone might be held to. The informality was the reason it was honest.

It was symmetrical. The engineer also picked up how the leader was doing, what the wider mood was, whether their work mattered. Information moved both ways.

Distribution removed all four at once. And it's worth noting this is largely invisible to the leaders it affects: they didn't lose a report, they lost a background sense they never consciously catalogued.

The uncomfortable version: it was always a kludge

Everything above treats corridor awareness as a good channel that distribution took away. There's a harder reading, and it's probably the more useful one.

The ad-hoc conversation was never a process. It wasn't repeatable and it didn't scale — it worked because you happened to be in the same building as the people you needed information from, and it stopped working the moment anyone was on holiday, off-site, or one floor away. Its failures were always there; they were just attributed to bad luck rather than to the mechanism.

Every time a team was surprised by a missed deadline in an office, the same gap existed. Distribution didn't create the problem. It removed the thing that had been papering over it, which is why so many organisations discovered on going remote that their status reporting had never really worked.

That reframing matters practically: if you treat the corridor as the gold standard, the goal becomes simulating it, and you end up scheduling calls to recreate accidents of proximity. If you treat it as a crutch, the goal is a mechanism that would have been better than the corridor even in an office — explicit, regular, and available to everyone rather than to whoever was nearby.

Why more meetings is the wrong fix

The instinctive response is to schedule the contact that used to happen by itself. More one-to-ones, a daily leadership sync, a weekly project review. One principal project manager described where this ends: a thirty-minute scrum-of-scrums at nine every morning, daily leadership standups with all PMs and team leads, and written updates on top. His verdict on the arrangement:

"Keeping abreast of status: a lot of effort and a pain in the ass."

Three reasons it doesn't reproduce what was lost.

It's expensive, so it's infrequent. A corridor conversation costs nothing and happens daily. A meeting costs an hour of several people's time and happens weekly. You've replaced a continuous signal with a sampled one, and the sampling interval is where problems hide.

It changes what gets said. A scheduled call with your skip-level is a performance in a way that passing them on the stairs isn't. People prepare. Preparation means deciding what to raise, which means deciding what not to raise.

Time zones make it worse. Every synchronous ritual across a spread of time zones is paid for by someone taking a call at seven in the morning or nine at night, and the tax falls on the same people repeatedly.

What a replacement has to provide

Working from the four properties:

• Continuous, or near enough. Something that happens weekly or daily rather than at review boundaries.
• Nearly free. If it costs a person more than a few seconds, it won't run at a useful frequency — and it'll be resented at any frequency.
• Low-stakes. Registering mild doubt has to be cheap. This is where anonymity does real work: it recreates the off-the-record quality that made corridor conversation honest.
• Comparable over time. Continuous presence gave leaders a baseline without anyone recording anything. Distributed, the only substitute is an actual record.

That last point is the one most often missed. The office version worked because you were there for the change — you noticed the team getting quieter because you'd been present for the noisier version. Remotely, nobody carries that baseline, so it has to exist in a system or not at all.

What to keep synchronous

Not everything should move to async, and it's worth being clear about what shouldn't.

Unblocking. When someone is stuck, a conversation resolves it faster than any amount of writing. Keep the part of standup that does this.

Diagnosis. A signal tells you something is wrong; it never tells you what. That conversation is synchronous and should be.

The relationship. One-to-ones do work no measurement reaches. What changes is that they stop being spent on status recitation.

The principle: use asynchronous mechanisms for detection and synchronous time for resolution. Most distributed teams have it backwards, spending synchronous time detecting things a five-second signal would have surfaced, and then handling the resolution over Slack threads.

The hybrid trap

One specific failure worth flagging. In a hybrid team, the ambient channel still works — for the people in the office. The leader who's in three days a week has a rich sense of how the co-located half is doing and a much thinner one of everyone else, without necessarily noticing the asymmetry.

The practical consequence is that remote team members get less attention when things go wrong, and their concerns surface later. If you're running hybrid, the asynchronous signal matters more, not less — because it's the only channel that treats everyone the same.

How Genchi does this

Genchi asks every team member the same question at a cadence you choose, delivered in Slack, email or the app: how confident are you that we'll achieve our goal? One to five, about two seconds.

That maps onto the four properties. It runs continuously rather than at review boundaries. It's near-free for the respondent. Responses are aggregated before anyone sees them, which restores the off-the-record quality corridor conversation used to provide. And because it's recorded, the trend supplies the baseline that being physically present used to supply for free.