Status Is Green Because Green Is Safe

The slide everyone agrees to believe

Picture the steering review for a six-hundred-million-dollar transformation, eighteen months in. The wall is a field of green dots, one per initiative, and a sponsor walks the room through them at one every forty seconds. Everyone nods. What nobody says out loud is that three of those green dots are known to be red by the people closest to them, some of whom are in the room. The owner of the biggest workstream believes his number is thirty percent high and has since the target was handed down. The controller netted two workstreams claiming the same saving last week and settled the credit by which sponsor was louder, without recording that she had a choice. At two sites marked complete, the old spreadsheet still runs the busy hours. None of it is on the wall. The wall is green, the program asks for its next funding wave, and the wall wins.

Change managers have a name for this that never made it into a textbook: the watermelon project. Green outside, red inside. It gets treated as folklore, a thing operators warn each other about over drinks. It is not. It is the most reliably documented failure mode in the literature on running large programs, and the mechanism behind it is understood well enough that the surprise is how rarely the people approving the money have heard of it.

Why bad news does not climb

Start with the mechanism, because it is the whole story. Mark Keil and colleagues, across fourteen studies of project status reporting^[1], established what they call the mum effect: people reliably withhold bad news as it travels upward, even against their own interest, even when they are the auditor whose job is to report it. The reluctance is not laziness or malice but a rational read of the local incentive. The first to color a status red owns the conversation about why and a reputation for bringing problems. Saying nothing costs nothing today; the cost shows up later, on someone else's watch.

Now stack that incentive at every layer and watch a single true fact climb. An owner who knows his target is unreachable softens red to amber, because the first honest red gets punished and amber buys a month. His manager rounds the amber up. The transformation office averages the workstreams into a portfolio color, and the portfolio is green. The board receives the green. Nobody lied. Each did the locally sensible thing, which at every step was to shave a little optimism onto the number, and the sum is a structural guarantee: the decision-maker is the last in the building to learn what the floor knew first. The signal does not survive the climb. It was never going to.

The loss is worse than most sponsors assume, because the optimism compounds before the program even starts. Michael Mankins and Richard Steele surveyed senior executives at 197 large companies^[2] and found that strategies deliver, on average, only sixty-three percent of their promised financial value, the largest named leak being inadequate or unavailable resources. Their sharpest observation is the cultural one: when plans routinely miss, leadership loses the ability to tell whether the gap is bad strategy, bad execution, or both, and unrealistic plans quietly become expected. The dashboard becomes theater that all parties perform and none believe.

The number the floor cannot name

If the status is fiction, you might hope the strategy underneath is understood. It usually is not. Donald Sull, with Rebecca Homkes and Charles Sull, surveyed roughly 7,600 managers across 262 companies^[3] over nearly a decade, and asked the simplest question: name your company's top strategic priorities in your own words. Fewer than a third of senior executives' direct reports could clearly connect to them, collapsing to sixteen percent for frontline supervisors, and fifty-five percent of middle managers could not name even one of the top five. In one firm whose own engagement survey reported eighty-four percent of staff clear on priorities, fewer than a third of the top team could name even two.

Sull's deeper finding reframes the rescue problem. Execution does not fail along the vertical chain management instinctively tightens. Eighty-four percent of managers can rely on their boss and their own reports; only nine percent can rely on colleagues in other functions all the time, barely better than they trust external partners. The execution problem is horizontal, living in the seams where no system has teeth. The trap, which Sull names directly, is that a leadership team watching the program stall responds with more metrics, more meetings, and tighter top-down alignment, the exact lever that makes a horizontal failure worse. It tightens the chain that works and starves the seam that does not.

And almost nothing dies. Scott Anthony and colleagues at Innosight^[4] found that in one IT company's audit, roughly twenty percent of the active portfolio were zombies, projects shuffling forward with no real hope, kept alive because killing one means telling a colleague their work was wasted and maybe costing them headcount. A zombie does not just burn its own budget; it holds the scarce experts whose names appear on three charters at once away from work that could actually pay. The portfolio looks busy and governed, and much of the motion is undead.

What the gap actually costs

Put the misreporting, the priority fog, and the zombies together and the leak is not a tail risk; it is the base rate. McKinsey's transformation research^[5] found that even self-declared successful transformations capture only sixty-seven percent of the maximum financial benefit, and everyone else captures thirty-seven, with roughly a quarter of the loss locked in before the work begins, at target setting, where the number is negotiated rather than measured. Run that against a six-hundred-million-dollar program at the average and the shortfall runs well over a hundred million dollars the green dots never showed.

The damning part is that they do not move. Bain's analysis of business transformations^[6] puts the share that produce lasting results at twelve percent, flat for two decades, with eighty-eight percent failing their original ambitions. Bain's own conclusion points back at the room: roughly eighty percent of the risk is internal and predictable, sitting in sponsorship, decision speed, and the overload of the same top performers across too many initiatives. What kills the program is rarely the market but the organization's own structure, which it can usually see.

That claim has its own evidence. Matthew Olson and Derek van Bever, studying growth stalls^[7], found that eighty-seven percent of large companies stalled at least once, that the average loser shed seventy-four percent of its market capitalization relative to the index, and that only thirteen percent of stall causes were external. The rest were management choices, found by an assumptions audit: write down the load-bearing beliefs the strategy rests on and test which quietly expired. Their refrain, case after case, is that insiders usually knew, and the signal sat in the heads of people who had watched the strategy die with no safe channel to say so.

One honest counterweight: the rescue trap is assuming every red is recoverable information. BCG's DICE research, from a 225-project study applied to more than a thousand programs^[8], scores the hard factors a program can control. Sometimes the floor is telling the truth and the operating model is genuinely broken, and listening harder will not staff the shift. But the canonical DICE failure is the inverse of denial: a CEO told BCG he was doing more than enough to back a project while his line managers said he had extended almost none. Same program, two irreconcilable readings, and the gap was the diagnosis. Nobody had put the same question to both and compared.

Why the truth stays in the building

Pull the threads together and they converge on one uncomfortable shape. The information that would rescue a stalled program is not missing. It is distributed across the people doing the work, and it stays put because every channel built to collect it punishes honesty. The owner who flags his real number hands in early notice on his own credibility; the controller who writes down a double-counted benefit starts a fight with a sponsor. So each stays quiet, and the steering deck rolls it into a color that reaches the board scrubbed of everything that mattered. Doug Yakola, who runs turnarounds from inside distressed companies^[9], calls the slow version of this managers boiling like frogs, backing into a crisis each half-saw coming. Every silence is locally rational. The aggregate is a decision-maker on instruments calibrated to read clear no matter the weather.

The grim symmetry is that every expert above describes the same recovery move. Sull compares priorities named in people's own words across layers. DICE reads the spread when executives and the floor score the same project separately. Keil's answer to the mum effect is independent reporting lines and making red a safe color. The instinct, when a program stalls, is to demand sharper reporting, and that is the trap: sharper reporting through the same channels produces only a cleaner version of the same lie. The truth is not unknowable. The only instruments built for asking are the ones the floor learned long ago to report clear into.

Which points, finally, at the kind of instrument this calls for. Not another status dashboard, which aggregates the softened number by design. Not another engagement survey, which the firm above used to report eighty-four percent clarity over a leadership team that could not name two priorities. Something closer to what the canon has described for years and few have run at scale: a neutral, confidential conversation anchored to a specific recent week, asking the owner and the controller and the supervisor what they actually did and what they left out, with one layer's answers set against another to find the contradiction. That is the instrument we are building at Latent Variables. The diagnosis was settled long ago; the open problem was only ever reaching the people who hold it before the next funding wave clears the green wall.