How do I track 20 teams without micromanaging?

The bind at scale

A VP of engineering responsible for 120 engineers described the everyday experience:

"It is difficult — stressful, tiring — to stay across the current status of over a dozen projects."

And the specific moment that makes it stressful:

"I hate seeing [an executive] in the corridor and being asked: what is going on with Project X?"

That's not vanity. Not knowing means you can't answer, can't reassure, and can't act. But the available responses to not knowing are both bad.

Sample. Hold one-to-ones, attend some reviews, read the reports. This is what most leaders do, and its flaw is selection: you hear from the people you meet most often, the ones comfortable raising things, and the teams whose leads are good at summarising. The quiet team with a problem is precisely the one you don't hear from. A group project manager described the consequence of operating one step removed:

"The problem with being one step removed — if the PM had detected the thing sooner, it would have been helpful."

Go deep. Attend standups, read tickets, ask direct questions of engineers. This produces better information about a small number of teams and costs you the ability to do your job. It also has a cost you may not see, described by a head of engineering:

"Me jumping on an issue is threatening to engineers."

Depth from a senior leader isn't experienced as interest. It's experienced as scrutiny, and teams adapt by presenting a smoother picture — which degrades the very information you went looking for.

What leaders did before: managing by walking around

The traditional answer to this problem wasn't a report. It was presence — walking the halls, an open-door policy, conversations at the water cooler. The VP of engineering quoted above listed "walking the halls" alongside one-to-ones and Project Central tickets as one of the three ways he actually knew what was going on.

He was explicit about what he was hunting for:

"Looking for the cracks in the paint. Little smouldering fires."

"Needs to know what fires are smouldering."

That's the right description of the job. Not finding out which projects are on fire — those announce themselves — but noticing the ones beginning to smoulder, while a bucket of water would still be enough.

Managing by walking around worked because it was continuous, cost nothing, and picked up things nobody would put in writing: tone of voice, who looked tired, an offhand remark on the way back from lunch. Its limits are equally clear. It doesn't scale past what one person can physically cover, it samples whoever happens to be at their desk, and it disappeared entirely when teams distributed.

The same leader described what he actually wanted from any replacement, and it's a good specification:

"Don't care about the words — is the product in trouble or not? Do I need to double click or not?"

The distinction that resolves it

The reason these feel like the only options is that in most organisations, the only way to find out is to ask — and asking is itself an intervention. A senior leader's question is never just a question. It signals attention, and attention signals concern.

If awareness can be made passive — arriving without anyone being asked, at no cost to anyone — the two come apart. You can know without anyone knowing you looked. Intervention becomes a decision you make afterwards, rather than the price of finding out.

That's the whole argument for a portfolio signal, and it's why the objection "isn't this just surveillance at scale?" gets the answer backwards. The intrusive version is the one where you have to go and ask.

What you actually do with 20 signals

Being able to see twenty teams doesn't mean thinking about twenty teams. The practical use is subtractive.

Most of your attention goes nowhere. Sixteen of the twenty are steady. You look, you register that, you move on. That takes about fifteen seconds and it's the majority of the value — not because you learned anything, but because you can now stop wondering.

A leader in our research described the failure this prevents:

"We waste a lot of cycles on projects that are in the green, but we want to talk about the red projects."

Status meetings distribute attention evenly by design — everyone reports, healthy or not. A portfolio view distributes it by need.

A few get a second look. Two or three showing drift. You look at the trend, read any comments, and decide whether to do anything — which is frequently no, or not yet.

Occasionally one gets a conversation. And that conversation is with the team lead, not with the team. Almost always.

Three habits that keep it from becoming micromanagement

Route through the team lead by default. If your response to a dip is to appear in front of the team, you've converted a cheap signal into an expensive event. One conversation with one person is nearly always the right move.

Don't ask about the number. "Why is your team's confidence low?" puts a lead in the position of defending a figure they didn't individually produce, and the quickest way to make that question go away next month is a higher figure. Ask about the work: what's in the way, what would help.

Let visible dips pass without comment. Not every decline needs a response, and most don't — movement is normal, covered in what to do when the trend goes down. Teams watching you not panic at amber is what keeps the next signal honest.

The cross-team pattern you can only see at portfolio level

One thing that only becomes visible with breadth: correlated declines. If three teams that all depend on the same platform migration soften in the same fortnight, that's a signal about the migration, and none of the three teams can see it — each experiences a local problem.

A leader we interviewed identified this as the reason to want scale in the first place:

"From a scalability perspective — to look across the organisation and see where things are popping up."

How Genchi presents a portfolio

Genchi shows a colour-coded dashboard of every initiative in the organisation, on one screen. Not a sample, not the ones someone chose to escalate — all of them.

Each card is colour-coded from current input from every single team member on that initiative. The colour isn't a judgement someone entered; it's derived from what the people doing the work said this week. So the answer to "is this in trouble?" doesn't require you to guess, to ask, or to wait for someone to decide it's worth telling you.

Cards can be organised into whatever structure matches how your organisation actually works — by programme, by product area, by department, by any parent-child arrangement that makes sense. A programme made of six workstreams rolls up into one card you can open, rather than presenting as six unrelated tiles.

Double-click any card and you get the detail behind the colour: the current aggregated score, the spread of individual values so you can see whether the team agrees, the trend across recent check-ins, and any comments or blockers people raised. Which is the "do I need to double click or not?" question answered directly — the dashboard tells you where, and the detail tells you what.

What the portfolio deliberately doesn't contain is people. There's no per-person view, no participation tracking and no individual history — not hidden, not built. It's a map of work, not of workers.