Somewhere in your last planning cycle, your organization made a commitment that requires four functions to hold one position for three years. Most of those functions are measured quarterly.

Nobody in the room reconciled those two facts, and the commitment went into the plan anyway.

Commitment capacity is the property I find least examined at the executive level. It is the length of promise an organisation can make and then actually keep, and it gets set by the shortest incentive horizon among the functions the promise depends on.

An organisation whose engineering roadmap runs on annual planning, whose sales compensation resets quarterly, and whose supply chain is measured on current-quarter cost can sign a five-year agreement without difficulty.

Holding it is a separate question. Holding it requires each of those functions to absorb visible short-term pain in a quarter where their own measurement punishes them for absorbing it.

When commitment capacity runs low, the instinctive fix is coordination. A steering group, a cross-functional forum, a program office assembled to hold the functions together across the horizon.

Each addition is locally sensible, and each one moves the decision further from the people holding the information. The organisation accumulates more alignment activity and less capacity to commit, because commitment requires a named person to accept exposure, and coordination structures are designed to distribute exposure until nobody carries enough of it to be blamed.

That is the mechanism behind a pattern most executives have witnessed without naming. The company that talks about collaboration constantly is frequently the one that cannot make a decision stick, and the correlation runs in the direction people find counterintuitive.

SanDisk's exit from quarterly memory pricing is the clearest recent example of the opposite condition. The company moved cloud buyers onto multi-year supply agreements, three of which carry roughly $42 billion in minimum contractual revenue backed by more than $11 billion in enforceable guarantees.

Signing that required engineering to commit capacity years out, finance to underwrite the exposure, and sales to decline better spot pricing while sitting across from a counterparty specifically trained to find the seam between those three positions. Any one of them breaking would have collapsed the arrangement.

The commercial strategy was downstream of an organisational property. A company whose functions each optimise their own quarter cannot make that offer credibly, and no amount of stated cultural commitment to collaboration changes what the compensation plans are actually rewarding.

Take the longest commitment currently on your books and list every function it depends on. Then look at how each of those functions is measured and over what period.

Wherever the measurement horizon is shorter than the commitment, you have located the place the promise will break. The break will arrive as a surprise to everyone except the person whose compensation structure made it inevitable, and that person usually knew for months and had no incentive to raise it.

The executive teams I work with usually discover their commitment capacity by exceeding it, which is an expensive way to learn a number you could have calculated in an afternoon.

The calculation requires nothing more than an honest inventory of measurement horizons, and it tends to be uncomfortable for the same reason it is useful: the answer implicates the incentive design rather than the people operating inside it.

What your organisation can promise is a structural fact about how it is measured. It has very little to do with how much the leadership team wants to keep its word.

The longer version of this, including what I heard David Goeckeler say about getting out of the boom-bust cycle, is on Substack.