Every company has four or five people who actually know how the work gets done. Not the version in the process doc. The real one. You can probably name yours right now, and I’d bet none of them sit on your leadership team.

Remember them. They’re the reason this article exists.

Earlier this year I’m on a call with an ops leader who’s running four functions out of Smartsheet. There’s a portfolio view at the top that nobody trusts, because every team quietly keeps its own. She wrote the business case herself, which usually means IT had a longer queue than she was willing to sit in.

Twenty minutes in, the platform question is settled. monday.com. I didn’t sell her on it. She’d worked that out before we ever spoke.

Then she asks how long the move takes. And somewhere in the middle of my answer she goes quiet and starts adding something up.

It isn’t licensing. Whatever it is, it comes out to no.

She was counting her four or five.

Not headcount. Names. The people who know where the bodies are buried in a workflow that got built three reorgs ago. Usually not senior. Usually the one you call at eleven at night when something breaks.

Discovery, the way all of us run it, eats exactly those people for a quarter. Same people holding the operation together while it happens.

So the business case has one number on it and the real price is two. The second one never makes it onto a slide. There’s nowhere to put it.

We decide like it’s 2026 and deliver like it’s 2010

Look at what she just did. Four functions, replatformed, decided on a Tuesday afternoon. Meanwhile there’s a board somewhere approving an entire transformation agenda in a single meeting because a competitor moved first.

Big decisions are fast now. That’s the big potato.

Everything after it is small potatoes. What do we call this status. Does this approval step still apply. Whose naming convention wins when two teams disagree and both of them are certain they’re right.

I’m not mocking any of that. Get it wrong and nobody uses what you built. But that’s where the quarter goes.

And it doesn’t go there because the questions are hard. It goes there because we’ve decided the only way to get the answers is to ask. And asking means those four or five people. In rooms. For weeks.

The answers were already in the system she wanted to leave

The frustrating part is that nothing was missing.

Every decision that team ever made is sitting inside the platform they’re trying to get off. Every workaround. Every status that stopped meaning anything in 2023 and nobody deleted. It’s all in there, in the sheets, in the automations, in the columns nobody can explain anymore.

She didn’t need a better platform. She’d found one. She needed a way onto it that didn’t start by pulling her best people out of the business for three months.

That’s what we mean by Contextual Delivery. Read the environment instead of interviewing it. Then put something running in front of the people who know the work, and let them react.

They’re better at that anyway. Everyone is. Ask someone to describe how a process works and you get the tidy version. Show them a version that’s slightly wrong and they’ll correct you in four seconds flat.

I’ll be straight about where this stops, though. Everything after it still takes as long as it takes. A department head who’s run a process for four years has to accept that it’s changing, and nobody has figured out how to speed that up.

She wasn’t wrong to walk away, by the way. On the numbers in front of her, the move wasn’t worth what it cost.

Which means somewhere inside her company there’s a decision that was properly argued, properly funded, fully agreed, and sitting completely still. Waiting on a quarter that keeps moving.

Every enterprise has a few of those. Nobody counts them, because there’s no line item for strategy that got approved and never happened.

I’d like to know what that number looks like in yours.