The portfolio was full. The strategy wasn’t.

At a glance

Organization

Technology-enabled services enterprise, $2.2B revenue, multiple business units competing for a finite transformation budget.

Situation

74 initiatives in the proposed transformation portfolio, carrying $142M of claimed annual value, presented for funding approval.

Complication

The portfolio described what business units wanted to do. It did not establish what the strategy required the enterprise to be able to do — and the claimed value had never been reconciled across initiatives.

Entry point

Zero-Based Scoping™, before commitment. GET and SORT were the most visible stages of the lifecycle.

Result

$142M claimed → $78M underwritten → $70M committed, against which $51M of modeled first-year realization is measured. The funded portfolio narrowed to roughly 33 to 38 initiatives.

A portfolio assembled the way most portfolios are

The enterprise ran multiple business units against a single, finite pool of transformation funding. The portfolio had formed the way portfolios usually form: each unit proposed the initiatives it believed would move its own numbers, each proposal arrived with a sponsor and a benefit case, and the transformation office aggregated them into one view for approval.

By the time that view reached the executive committee it was substantial — 74 initiatives, $142M of claimed annual value, and a roadmap spanning several years. Nothing about it looked unreasonable. Every line had a name against it and a number beside it, and the total was large enough to feel like an answer to the strategy.

Management was not being careless. Bottom-up assembly is the normal way an enterprise finds out what its operators think needs to change, and it surfaces problems that no central team would see. The difficulty is what happens when that list is then treated as the transformation itself.

The problem was not a shortage of initiatives

It was the opposite. There was no shortage of things to do, no shortage of sponsorship, and no shortage of projected value. What was missing was evidence that the portfolio, taken as a whole, represented the operating changes required to deliver the strategy. Three gaps sat underneath that.

• Aggregation had never been tested. The $142M was a sum of independently constructed cases. No one had asked whether two initiatives were claiming the same benefit by different routes — and with business units proposing separately, several were.
• Attribution had never been challenged. Some cases assumed value that would arrive with or without the transformation: market movement, pricing actions already underway, decisions already taken elsewhere in the business.
• Requirements had never been derived. The portfolio described what the units wanted to do. It did not describe what the strategy required the enterprise to be able to do, so no one could see what was absent.

The third gap matters most and gets noticed least. A portfolio built from the bottom up will always look complete, because it is complete with respect to the proposals that were submitted. It says nothing at all about the requirements nobody proposed against.

Before we fund all of this, are these actually the right things to transform?

That was the question the executive committee could not answer from the material in front of it. It could confirm that the initiatives were credible individually. It could not confirm that they were the right set, that the total was real, or that anything essential was missing.

Why prioritization does not resolve it

The conventional response to a crowded portfolio is prioritization: score the initiatives, rank them, fund down the list until the money runs out. Prioritization is a sound discipline pointed at the wrong question here. It assumes the list is the right list and that the task is choosing among its members.

If the list was assembled from the bottom up, ranking it funds the best-argued proposals rather than the necessary ones. It also leaves duplicated benefit intact, because two overlapping cases will both score well — and the double count survives into the committed total.

Stage-gating has the same limitation from a different angle. A gate tests whether an initiative is ready to proceed. It does not test whether the initiative should exist, whether its benefit is already counted somewhere else, or whether the portfolio as a whole will produce the outcome the strategy needs. Both disciplines govern the members of the set. Neither governs the set.

The approach: start from the outcome, not the list

Zero-Based Transformation refuses the inherited baseline. In portfolio formation that means the starting point is not the 74 initiatives. It is the outcomes the transformation has to produce, and the operating changes required to produce them. The portfolio is then tested against that, rather than being ranked against itself.

First, the outcomes the transformation actually had to produce

Stated in business terms and owned by named executives — not as strategic themes, which every initiative can be mapped to, but as specific enterprise results with a number attached. This is deliberately done before looking at the initiative list, so the list cannot shape the definition of success.

Second, the operating changes those outcomes require

Working backward from each outcome to what has to change in how the enterprise actually runs: how work enters, how it is classified and priced, how it is executed and handed off, how completion is evidenced, and how the model corrects itself. This produces a requirement set the enterprise did not have, expressed independently of anyone’s proposal.

Third, the test against the proposed portfolio

Each initiative was examined for the relationship it claimed to an outcome. Overlapping value claims were surfaced by reading the initiatives against one another rather than in isolation. Activity was separated from required operating change — a distinction that removes a surprising amount of a typical portfolio, because much of what gets proposed improves how a function works without changing what the enterprise can do.

Fourth, a higher evidentiary standard before commitment

Value was not accepted because a sponsor had asserted it. Each claim had to survive on attribution and on the strength of the path that would deliver it. Where it could not, the value was adjusted or removed — before the money was committed, rather than discovered eighteen months later in a benefits review.

Where an outcome had a requirement with no funded initiative behind it, that gap was named and priced. It is the part of the exercise leadership tends to value most, because bottom-up portfolios are systematically silent about what is missing.

The reconciliation: $142M to $70M

Three categories of adjustment account for the movement between the claimed total and the committed one.

Overlapping claims — $28M

Business units proposing independently had, in several cases, built benefit cases on the same underlying improvement. Read side by side rather than sequentially, the double count was visible. None of it was dishonest; it is a structural consequence of aggregating separately constructed cases.

Attribution — $18M

A benefit case can be arithmetically correct and still not belong to the transformation. Where value was going to arrive regardless, or where the causal link to a specific operating change could not be established, it was removed from the transformation’s account. It may still arrive. It should not be used to justify transformation spend.

Delivery and evidence path — $18M

The remaining adjustment was neither duplicated nor misattributed. It was value whose delivery path was not strong enough to support the amount claimed — dependencies unresolved, ownership unclear, or no route by which the result could later be evidenced. These initiatives were largely retained at an adjusted value rather than canceled.

The point is not that half the business case was destroyed. It is that a headline number became a commitment — one leadership could govern, report and defend.

What was decided

The funded portfolio narrowed from 74 initiatives to roughly 33 to 38. Some were retired outright, some consolidated where two initiatives were pursuing one operating change, and some retained at reduced value with a clearer requirement attached.

The model output was $78M. Leadership committed to $70M. That $8M gap is not a modeled probability or a risk-weighted haircut — it is a management decision to hold value outside the committed case and be measured against the lower figure. Keeping the distinction visible is the point: an underwritten number is what the evidence supports, and a committed number is what leadership chooses to be held to.

Requirements with no initiative behind them were funded. That is the part of the decision that does not show in the arithmetic, and the part most likely to determine whether the strategy converts.

Most importantly, funding moved from a list of proposed projects to a portfolio explicitly tied to the business outcomes it was required to produce. The governance conversation changed with it: from which initiatives are on track, to how much of the committed value has arrived.

The results

Against the $70M commitment, modeled first-year realization is $51M — 72.9% of committed value. Measured against the original $142M headline, the same $51M would read as 35.9%, which is precisely why the headline was the wrong number to govern against. The $51M is a modeled realization figure for this case, derived against the $70M commitment; it is not fitted to any external benchmark, and the resemblance to published first-year realization rates is incidental.

The gross-to-committed reduction was 50.7%. Read as a headline that sounds like value destruction. Read commercially it is the opposite: leadership stopped carrying $72M of value it could not have defended, and started reporting against a number that survived scrutiny before anyone was accountable for it.

The reduction also arrived at the only point where it is cheap. Removing $28M of duplicated benefit before funding costs a difficult conversation. Discovering it during a benefits review costs the credibility of the entire program.

Why the result held

Realization tracks because the case had already been tested. Value that could not be attributed was never committed, so there is nothing to unwind when the commitment is put to Finance validation — which is the proof mechanism the model applies at each reporting period. The initiatives that survived carried a stated requirement and a named owner, which meant progress could be assessed against something other than schedule.

The commitment decision did the rest. Because leadership had chosen to be measured against $70M rather than the $78M the model supported, the first delivery problem did not force a restatement of the committed total. That headroom was a governance choice, not a reserve mechanism, and it was visible as such to everyone reporting against it.

What this means if you are at the same point

The pattern in this case is common enough to be worth testing directly, whatever the size of the portfolio.

If you cannot say, from the material in front of you, which operating changes your strategy requires and which of them are funded, the portfolio has not yet been tested — regardless of how many initiatives it contains.

Evidence basis

The situation this case describes is well documented. A material share of transformation value is decided at target setting, before delivery is meaningfully underway. The findings below establish that this pattern and these economics are plausible; they are cited as context and are not the source of the case figures.

• McKinsey, Losing from day one: why even successful transformations fall short (December 2021 global survey, n=1,034). Respondents report that nearly one-quarter of transformation value loss occurs during the target-setting phase, before implementation is well underway, and that 55% occurs during and after implementation.
• McKinsey, The numbers behind successful transformations — an analysis of 82 publicly listed companies with observable 18-month transformation records, distinct from the survey above. Successful transformations implemented initiatives corresponding to 74% of fully ramped value within the first 12 months. Cited as context only; it was not used to set this case’s modeled realization figure.