Over four decades of managing IT and business process organizations, from the World Bank to Wall Street to a Fortune 200 insurer, I have noticed something that surprises almost every executive I coach: the teams that ship the most projects are not always the teams that create the most value.

That sounds like heresy in a culture obsessed with velocity and delivery dashboards. But delivery is a means to an end. The end is achieving, and sustaining, the business outcomes senior management has defined and prioritized. If your organization cannot articulate that distinction clearly, no amount of process will save you.

Here is how I think about building governance that actually works, not the kind that suffocates an organization, but the kind that keeps hard-won budget savings from quietly disappearing.

Governance starts with outcomes, not paperwork

Before you design a single approval workflow, senior management needs to define and communicate the business outcomes it is prioritizing. Skip this step and your governance process becomes a box-checking exercise: people learn to write proposals that get approved, not proposals that create value.

Keep the threshold reasonable, and the documentation standard

I have always recommended a simple rule: any project above a defined size, say $250K or six person-months of effort, requires a standard package. A brief description. The specific business outcomes it will achieve. The team. The timeline. And a real budget, broken into capex, one-time opex, and ongoing opex.

That last distinction matters more than people think. At the World Bank, I once asked our budgeting director why we only capitalized contract labor when accounting rules allow staff labor to be capitalized too, provided the work qualifies. The next fiscal year, we changed the policy and it materially changed how every team in the division thought about its budget. Small governance decisions like this ripple through an entire organization.

Separate IT governance from business governance

I recommend a carefully crafted agenda when combining IT and business representatives into a single governance meeting. There will be legitimate debates about technology strategy, sequencing, and trade-offs. Airing those disagreements in front of business partners undermines confidence in the whole department. Let IT leadership govern internal IT investments against a centralized budget. Let business and IT jointly govern the projects the business is funding, but resolve internal IT disagreements before that meeting, not during it.

Two governance models that actually worked

At the World Bank, we ran a disciplined, capex-driven process, paired with a Rapid Application Development team that handled anything under $100K or four person-months without going through full governance. The smaller RAD project queue stayed visible in real time to every governance member in the event that they wanted to weigh in on the project. That gave our business partners speed on small asks without sacrificing oversight on the large ones.

On Wall Street, every major desk had a Managing Director — we affectionately called them "The Three Kings" — with deep business and technology backgrounds, who personally reviewed any project over $250K. Anything over $1M needed business leadership sign-off as well. Rejections were common, and it often took two or three revisions before a proposal was approved. That friction was the point. Ensuring business alignment of IT spend is a prerequisite to efficient operations.

Start light. Refine only where it hurts.

If I have one piece of advice for a leader building governance from scratch, it is this: start with a lightweight process, watch for the specific places where things go wrong, and add rigor only there. Overbuilt governance kills the agility that makes an organization competitive. Underbuilt governance lets institutional inertia quietly rebuild all the waste you just spent a year eliminating.

Because that is really what governance is for. It is not about control for its own sake. It is the discipline that keeps a cleaned-up budget clean — long after the excitement of the optimization project has faded and everyone's attention has moved to the next priority.

Continue reading: You Are Wasting 10 to 45 Percent of Your IT and BPO Budget · Vendor Management: The Least Painful Path to Significant Savings · Strategic Advisory Services