Annual Project Funding Against 38% Hidden Demand
Northwind Retail funds IT through annual project business cases approved every October, replanned quarterly by a PMO that reports 84% on-time delivery - achieved largely by re-baselining scope. Six value streams deliver across 42 applications, but the top three initiatives all depend on the same four senior engineers. Lead time from approved idea to first commit averages 97 days. Roughly 38% of delivery capacity is consumed by unplanned work (regulatory fixes, incidents, data corrections) that never appears in the plan, so every quarter is over-committed on paper. Prioritisation is a single ROI spreadsheet ranked by finance; the sales team can escalate any request into the next sprint and does so about nine times a quarter. Product owners see unplanned demand only when it lands on them. A new CIO wants continuous funding and flow-based planning, and asks you to design the change without losing the financial controls the board relies on.
1. Diagnose the planning failure modes. Which metrics expose what the 84% figure hides?
Read the worked answer
The 84% measures the PMO's ability to redefine success, not delivery performance. Three failure modes sit underneath it.
- Hidden demand. If 38% of capacity goes to unplanned work, the plan is at least 38% fiction. On-time delivery is computed against a baseline that excludes a third of the work, so the number is arithmetically guaranteed to look healthy.
- Contention dressed as parallelism. Three initiatives sharing four engineers means two are always waiting. The 97-day approval-to-first-commit figure is queue time, not work time. Little's Law says the fix is limiting work in progress, not adding projects to the queue.
- Prioritisation without a single queue. One ROI ranking plus unlimited escalation rights means priority is set by whoever escalates hardest. Nine escalations a quarter is small in absolute terms, but each one displaces planned work and restarts sizing conversations.
Metrics that expose it: end-to-end lead time (approved to in production), WIP per team and per constrained role, flow efficiency (active time divided by lead time), percentage of capacity on unplanned work reported as a first-class number, and escalation rate per value stream.
The metric to retire is on-time delivery against a movable baseline. Replace it with committed work completed inside the committed window, plus forecast accuracy comparing planned and actual capacity consumed.
2. Redesign prioritisation and funding for the next annual cycle. Be specific about what replaces the annual business case and the ROI spreadsheet.
Read the worked answer
Separate the two things the board actually wants: control of spend, and confidence about outcomes. Continuous funding does not mean no budget.
- Fund capability, not projects. Allocate a fixed quarterly budget envelope per value stream, sized from historic actuals including the 38% unplanned reality, owned by a named business sponsor. Initiatives become demand inside that envelope. The board still approves total spend and receives a quarterly outcome review.
- Ring-fence capacity for unplanned work. Reserve around 30-35% of each envelope explicitly. This is not slack; it stops planned work from being held accountable for work nobody planned.
- Replace the ROI spreadsheet with a WIP-limited ranked intake. One visible backlog per value stream, a monthly 60-minute prioritisation forum with a single named decision-maker, and published criteria (revenue, risk, cost, compliance deadline). Escalations route through the forum rather than around it, and a documented 'no' counts as a decision.
- Cap WIP at two initiatives in flight per value stream. New work enters only when one leaves.
- Make the four engineers visible as a constraint. Sequence the three initiatives or accept a longer lead time, and write down which you chose.
- Compare using cost of delay, not ROI alone. ROI spreadsheets hide time, which is precisely the variable hurting Northwind.
Main risk: an envelope with a weak sponsor becomes a slush fund. Attach the right to re-allocate at each quarterly review, and use it.
3. What should be preserved from the current model, and what are the risks of moving to continuous funding?
Read the worked answer
Preserve:
- The annual total spend approval and the audit trail from business case to spend. Finance is not the obstacle; losing financial traceability is how transformations get reversed in year two.
- Named business ownership of value streams, assuming the current sponsors are genuinely accountable rather than signatories.
- Regulatory and compliance deadlines as hard constraints in the ranking, never traded against ROI.
Risks of continuous funding:
- Envelope drift. Budgets quietly top up when a favoured initiative overruns. Without an explicit re-allocation rule this is project funding under a new name.
- Loss of the forcing function. The annual case makes someone justify benefits. If you remove it, put outcome reviews in place before removal, not after.
- Governance lag. Board control frameworks, auditors and capitalisation rules may still assume projects. Check with finance in month one, because retrofitting this later is expensive.
- A visible dip. The first two quarters of an envelope model often look worse while hidden demand surfaces and gets counted.
Pilot one value stream on envelopes for two quarters alongside the existing model, compare forecast accuracy and lead time, then extend. Switching all six at once converts a planning problem into a credibility problem.