Decision COE
The Tradeoff Has To Be Owned Before It Can Be Optimized
By Ashutosh Bansal·February 9, 2026
Every planning failure I have audited in the last five years has had the same root cause. Nobody named owned the tradeoff. So nobody could optimize it.
Every planning failure I have audited in the last five years has had the same root cause. It is not the forecast. It is not the data. It is not the platform. It is that nobody — no named individual, no chartered role, no governed body — owned the tradeoff.
Service versus margin. Margin versus working capital. Working capital versus resilience. These are the four corners of the operating cycle, and every consequential decision sits inside that quadrilateral. When the corners disagree, somebody has to call it. If no one is named, the call defaults to whoever escalates loudest in the meeting — usually the function with the most political weight that quarter.
This is not a planning problem. It is a governance problem. And it is the single most under-discussed reason transformation programs fail to deliver on their value case.
A platform — even an exceptional one — cannot fix this. The platform can present the tradeoff with full visibility, model the consequences with full transparency, and offer the optimal frontier with full mathematical defensibility. It cannot make the call. The call belongs to a human. The question is which human, and the answer cannot be “whoever is in the room.”
The Decision Center of Excellence is the structural answer. The COE charters named owners for every consequential tradeoff. The COE defines the escalation logic — what gets resolved at the policy level, what gets resolved at the COE level, what gets escalated to the executive committee, and under what conditions. The COE publishes the policy. The functions live with the policy. When the policy needs to change — because the business changed, because the volatility changed, because the strategy changed — the policy is revised, on cadence, in the open.
Without that structure, the platform is a participant in the same meeting that produced the failure. With the structure, the platform becomes the instrument by which the policy is enforced — and the policy becomes the instrument by which the tradeoff is owned.
The order matters. Own the tradeoff first. Optimize it second. Most enterprises try to do it the other way around. None of them succeed.
More from Insights
Decision Intelligence
Pilots That Prove the P&L, Not the Demo
There are two kinds of pilots in enterprise software. Only one of them tells you whether to write the next check.
January 28, 2026·A GitaCloud Principal
Enterprise Transformation
The 90-Day Roadmap That Survives the Boardroom
Most transformation roadmaps die in the first board review. The ones that survive share three structural properties — none of which are about the roadmap itself.
January 14, 2026·Ashutosh Bansal