Why your best ideas die in meetings and your worst projects survive in the exec committee

Two scenes, in any organization, in the same week.

First scene: an influential director defends his project in the exec committee. The return on investment is dubious, everyone senses it, but he has weight, a voice, a track record. The project goes through. It will consume budget and teams for months before anyone dares say it should never have started.

Second scene, three floors down: someone well-connected, who knows their turf inside out, has an idea worth its weight in gold. A simple adjustment, high return, low cost. They mention it in a team meeting. It's written down nowhere. It never travels up. The idea fades out, and no one will ever know it existed.

These two scenes are the same problem. In both cases, it wasn't value that decided, it was rank. A bad project survived because someone senior was backing it; a good idea died because no one important was backing it. Same bias, two directions.

Demand management is precisely the mechanism that corrects this — provided you understand what it really is.

Demand isn't a tollgate.

Demand management is the entry point for everything that could become a project in your organization. Every idea, every "we ought to do this," every request from a business unit passes through it — or should.

Most people reduce it to a filter: channel things, say no, keep people from launching any and everything. That's true, but it's half the job, and the less valuable half. Because a tollgate only stops. It's a defensive stance, and an organization that thinks of its demand solely as a checkpoint spends its time protecting itself from itself.

Its real role is twofold. On one side, it kills low-return projects early, before they devour budget — including when an influential sponsor pushes to keep them. On the other, and this is the hidden goldmine, it surfaces the high-potential good ideas that today die in silence, for lack of a channel or because they come from people with no political weight.

These two movements are really one. The principle that connects them fits in a single sentence: what rises to arbitration must be selected on its value, not on the rank of whoever is backing it. Everything else follows from that.

Kill early what rank protects.

Let's start with the side you think you've mastered: saying no. Except the real issue isn't saying no. It's saying no early, and saying it even when the person pushing outranks you.

The cost of a bad project isn't fixed. A low-return project killed in the first month costs almost nothing: a conversation, some framing, one arbitration. The same project killed in the ninth month has already mobilized teams, nibbled away at the budget, displaced other priorities, and will have become politically impossible to stop because too many people have poured energy into it. You don't kill a project a director has invested nine months of personal credit in. You let it live, and it keeps on costing.

Hence the importance of qualifying fast and qualifying the same way for everyone. A request backed by an influential director and a request backed by a project manager must be described the same way, tied to an expected value and to a capacity, laid on the same table. The day return on investment is judged on the evidence and not on the voice of whoever is speaking, the senior figure's bad project no longer goes through. Not because you've reined the director in, but because the system has stopped confusing conviction with value.

That's the first goldmine: the money you don't waste on projects that should never have existed, because you stopped them while they still cost nothing.

Save what the absence of rank condemns.

The other movement is quieter, and it's the more valuable of the two.

First, let's get rid of an illusion. People believe the hidden value lies in the big projects. It almost never does. The CRM switch, the ERP overhaul, the sweeping transformation program — those are the big rocks: they're visible on their own, they have a sponsor, they force their way into every arbitration. Nobody misses a big rock. Worse, they soak up all the attention and all the political energy available, and while you're debating the new CRM's schedule for the third time, no one is looking anywhere else.

Yet it's elsewhere that the needle hides in the haystack. The feature demanded by three regions at once. The process tweak that would save fifty people two days each. The high-return, low-cost idea coming up from the field. These things don't force their way in on their own. They have no sponsor. And that's exactly why they get lost.

What becomes of a good idea carried by someone with no title and no political weight? It dies. Not through a decision: no one sits down to say "this idea is bad." It's more insidious than that. There's no channel for it to rise through, it's never qualified, never compared with the others, and it fades out for lack of oxygen. The universal bias of organizations does the rest: what rises isn't what has the most value, it's what has the most voice. We politely call that hierarchy. In practice, it's a filter that selects rank and not value.

Here's the truth no dashboard will ever show you: your organization kills its best ideas without knowing it. And every nugget you let fade out is a gain that will never appear in the accounts, because you never count what you didn't do.

Listen to everything without drowning.

At this point the objection lands, and it's a fair one: "if I open the door to every idea from across the whole organization, I'll be swamped." True. And it's precisely this fear of volume that drives you to filter by default, to close the floodgates, to let through only what comes from above. You say no to the flood for fear of drowning, and in shutting the door on the noise, you shut it on the nuggets too.

This is where the setup changes everything. Capturing without drowning takes three things, and none of them is a technological miracle.

A single entry point, first. Every idea, wherever it comes from, enters through the same place, not through a buried form no one fills in. An identical qualification for everyone, next: every request described the same way, tied to a value and a capacity, so that it's judged on what it's worth and not on who's backing it. It's this standardization, however mundane, that neutralizes rank. Finally, something to absorb the noise. That's the role of artificial intelligence in AirSaas: it doesn't decide for you and it doesn't spot the needle on its own — it makes listening sustainable. It qualifies, groups together requests that look alike, links up the same idea surfaced from three different places, summarizes, and presents you with a clear landscape instead of an avalanche.

The leader keeps all their judgment. They only lose the noise. "Listening to everything" stops being a synonym for "putting up with everything," and you can finally open the door wide without anyone drowning.

Choosing on value is an act of leadership.

What a good CEO, a good CIO, a good head of transformation or operations really wants isn't to run projects. It's to be certain of two things at once: that not a single euro goes into a mediocre project just because someone senior defended it, and that no golden idea dies in some branch office just because no one important carried it.

That's what well-designed demand management is. Not a tollbooth at the entrance to the portfolio. A mechanism that selects value and refuses to confuse it with rank. It saves the money you would have wasted and captures the value you would have lost, and these two gains are the two faces of a single gesture. It often surfaces more than an idea, in fact: it reveals the person who carried it, that well-connected employee no org chart ever flagged.

An organization almost never lacks good ideas, nor the clear-sightedness to spot the bad ones. What it lacks is a system that judges on value rather than on sheer loudness. So the next time you sign off on your portfolio, ask both questions together: what are we funding today that doesn't deserve it, and what are we missing, right here, right now, for failure to have listened to it?

Rank knows how to make itself heard. Value, on the other hand, needs someone to go looking for it.

This article first appeared in my LinkedIn newsletter Il était une fois un CODIR, where every fortnight I tell the story of situations lived through in an executive committee.