The roadmap was laminated. That is the detail I remember most. 

Years ago, my leadership team closed a two-day planning offsite by approving 40 projects for the coming year. We had scored every candidate, debated the rankings, sequenced the winners into a roadmap, and printed it in color for the program wall. Budgets were allocated in full. People left aligned and proud. 

In the first week of March, I stood in front of that same roadmap with a marker, crossing projects out. Eleven of the 40 were stopped, rescoped, or changed beyond recognition. A reorganization had erased the business case for three of them. A regulatory change had jumped the queue and taken the exact team we had promised elsewhere. Two vendor slips had frozen a program for a quarter. 

None of it was bad planning in the usual sense. Every estimate had been defensible in January. 

The plan did not fail because we planned badly. It failed because we planned once. 

Why does the annual portfolio plan break by March? 

Look closely at how most annual portfolio planning cycles work and you find three design flaws baked in before the first project is scored. 

First, they are demand-first. The wish list is collected, scored, and ranked, and capacity is checked at the end, if at all. The result is a plan that consumes 100 percent or more of the organization's real delivery capacity, which means the first surprise of the year has nowhere to land. 

Second, they are funded in full. Every approved project receives its whole budget in January. From that moment, changing the plan means taking money away from someone, which turns every mid-year adjustment into a political fight. Leadership avoids the fight, so the plan stays frozen while reality moves. 

Third, they are decided at the point of maximum ignorance. January is when you know the least about the year ahead. Q1 actuals have not arrived. The reorganizations, regulatory changes, and market moves that will define the year have not happened yet. Locking your biggest resource decisions at that moment, permanently, is a habit we have all agreed not to question. 

A plan built this way becomes a museum exhibit. Accurate about the moment it was created, silent about everything that follows. 

Five disciplines for a 2027 plan that survives 

The fix is not better estimating or a longer offsite. It is building the plan, from the start, as an instrument that expects to change. Five disciplines do most of the work. 

1. Start with capacity, not the wish list 

Before scoring a single candidate project, measure your delivery capacity by role. Not headcount. Committed hours, by quarter, for the six or seven roles that actually constrain your portfolio: the solution architects, the change leads, the data engineers, whoever the real bottlenecks are. Then plan to no more than 85 percent of that number. The remaining margin is not slack. It is the landing zone for the surprises the year will send. A wish list scored into measured capacity produces a plan. A wish list scored against hope produces a document. 

2. Fund the year in tranches 

Fully fund the work starting in the first quarter. Hold the rest of the budget in a portfolio reserve, released at quarterly checkpoints only if the initiative still deserves it. This one change converts your budget from an entitlement into a steering instrument. Projects come to checkpoints with evidence. Stopping something is no longer a confiscation, it is a reserve that was never released. And new opportunities mid-year have a real funding source instead of requiring surgery on someone else's allocation. 

3. Put decision dates on the roadmap 

Delivery dates tell you when work finishes. Decision dates tell you when the plan itself is re-examined. Put them on the roadmap before January: the quarterly moments when leadership re-votes the remaining year against current facts. When those dates are on the calendar from the start, replanning is routine governance. When they are not, replanning only happens through crisis, and it arrives with blame attached. 

4. Publish the unfunded list 

Keep every scored-but-unfunded candidate on a visible list, with the reasons. This list is the plan's shock absorber. When new demand arrives in May, it competes against the unfunded list, not against a blank page, and the conversation immediately has context: is this better than the best thing we already said no to? The list also protects the portfolio from side doors. When executives can see what was deliberately deferred, projects find it much harder to appear from nowhere. 

5. Pre-agree the change rules 

Write into the plan itself what triggers a replan between decision dates, who decides, in what forum, and on what clock. A capacity shift beyond a threshold. The failure of a named business case assumption. New demand above a certain size. When the rules exist before the event, mid-year change is administration. When they do not, every change is a negotiation, and the loudest voice wins. 

Where Smartsheet makes this real 

Every one of these disciplines needs a live data spine, and this is where I lean on Smartsheet in client work. A capacity sheet holds committed hours by role and quarter, fed from real project schedules. One intake pipeline scores every 2027 candidate against strategy and against that capacity sheet at the same time. The portfolio roadmap carries decision-date milestones alongside delivery milestones. Control Center rolls tranche budgets up into a single executive dashboard where demand, capacity, and reserve are visible in one view. The platform does not make the decisions. It makes the trade-offs impossible to ignore while the decision is being made. 

What changed when the plan could bend 

The cycle after our laminated-roadmap year, we ran all five disciplines. The plan was challenged in February, adjusted in April, and rebalanced in July, and that is exactly the point. By October, the portfolio still recognizably served the strategy we had set, and the replanning that used to consume weeks of defensive slide-building took an afternoon per quarter. Leadership stopped asking whether the plan was right. They started asking whether the steering was working. That is a far better conversation for a PMO to host. 

Your one action for next week 

Pull the portfolio plan your organization approved in January 2026. Count the funded items. Count how many are still running roughly as scoped. Divide the second number by the first. That is your plan survival rate, and it is the single best argument for changing how your 2027 cycle works. Bring it to your next leadership meeting and ask two questions: what should this number be next year, and which of the five disciplines would move it most? 

If you want an experienced outside eye on your planning process before the 2027 cycle locks, this is exactly what the PMO Value Blueprint was built for: a focused 4 to 8 week assessment of how your portfolio is defined, funded, and steered, with an improvement roadmap you can act on before January. Book a conversation at pmoevolution.com/contact. No sales pitch, a 30-minute conversation. 

Visit our website: pmoevolution.com 
Follow us on LinkedIn: PMO Evolution 
Subscribe on YouTube: PMO Evolution