Screening a portfolio for value-study candidates

Selection is the first value decision

At Level 1 you learned to run a value study on a subject someone else had chosen. At practitioner level the question moves upstream: which subjects deserve a study at all, what kind of study, and when? This choice is itself a value decision. A study is not free. It consumes the time of a sponsor, a facilitator and a team of experienced people who are usually busy on the very projects being studied. A study pointed at the wrong subject, or at the right subject too late, spends that effort and returns little.

Most organizations do not have a shortage of candidates. A capital programme, a technology roadmap or an operations improvement plan normally holds far more work than any value office can examine. In practice the value practitioner is often asked, informally, to "look at the big ones". Size matters, but it is not enough. A very large project that is already in construction offers little room to change anything that matters. A modest project that sets a standard design for twenty future buildings may repay a study many times over.

Selection also protects the credibility of the value function. Studies that produce visible, implemented improvements build demand for more studies. Studies imposed on projects that cannot respond, because contracts are signed or designs are frozen, produce long lists of rejected proposals and the reputation that value studies are an obstacle. Choosing well is the practitioner's first contribution to the portfolio.

Finally, selection is where the value practitioner meets portfolio governance. Portfolio boards rank projects by strategic fit, risk and funding. A value office that uses compatible language, and can explain why a study on project A matters more to the strategy than one on project B, earns a seat in those conversations.

Signals that a subject will repay a study

Experienced practitioners look for a cluster of signals rather than a single trigger. The most useful are:

• Cost magnitude and life-cycle significance. Large capital cost is an obvious signal, but so is a large running cost. An energy plant with a modest capital cost may consume far more money in fuel and maintenance over 25 years than it cost to build.
• Remaining influence. How much of the design, scope or operating model is still open? A subject at concept stage, before key layout, technology and procurement decisions, offers far more room than one at tender.
• Complexity and uncertainty. Many interacting systems, unclear requirements, competing stakeholders or new technology all increase the chance that the current solution is not the best value.
• Strategic importance. Subjects that directly carry an organizational objective, such as capacity, safety, carbon or service access, deserve closer attention because the value at stake goes beyond cost.
• Problem signals. An estimate that exceeds the budget, a schedule that keeps slipping, stakeholders who disagree about what the project is for, or a design that has been redrawn several times all suggest unresolved function questions.
• A workable window. A study needs a few weeks of preparation and a point in the schedule where its results can be absorbed. A subject with a decision gate next week may need a different intervention.

Notice what is missing: "the project is in trouble" is not, by itself, a reason for a value study. Some troubled projects need governance, leadership or commercial action first. A value study works on function and alternatives, and it needs a sponsor willing to act on the results.

The PVology Academy Study Selection Screen

This is a PVology Academy framework. It turns the signals above into a transparent weighted score so that candidates can be compared on the same basis and the reasoning can be challenged. It is a teaching and discussion aid, not a formula that makes the decision.

The screen uses six criteria, each scored from 1 (weak signal) to 5 (strong signal):

1. C1 Cost and life-cycle significance, weight 25
2. C2 Remaining influence (design or decision stage), weight 20
3. C3 Complexity and uncertainty, weight 15
4. C4 Strategic importance, weight 20
5. C5 Problem signals, weight 10
6. C6 Workable study window, weight 10

The weights add to 100. The weighted score is the sum of weight multiplied by score, divided by 5, so a candidate scoring 5 on every criterion reaches 100 and one scoring 1 on every criterion reaches 20.

Weighted score = (sum of weight x score) / 5

Two rules keep the screen honest. First, write a one-line reason for every score. "C2 = 2 because the design is at 60% and the structural frame is fixed" is useful; a bare number is not. Second, score with at least two people, typically the value practitioner and a portfolio or finance representative, and discuss differences of more than one point. The conversation is often more valuable than the number.

Organizations should tailor the weights. A public health authority with a carbon commitment may raise life-cycle significance; a technology company shipping a product every quarter may raise the study window weight. Whatever the weights, publish them before scoring begins, so that nobody adjusts them to favour a preferred project.

Matching the type of study to the candidate

The screen ranks candidates, but it does not say that every high-scoring candidate needs the same intervention. Practitioners choose among several study types:

• Full value study. All eight Job Plan phases, a multidisciplinary team, several workshop days and a formal implementation phase. Suited to large, complex subjects with high remaining influence.
• Focused workshop. One or two days on a defined part of the subject, such as a building's mechanical systems or a software platform's integration layer. Suited to subjects where one or two functions dominate cost or risk.
• Function and requirements review. A shorter session at the very start of a project that tests whether the stated need, functions and performance requirements are right. Suited to early-stage subjects where design has not started.
• Design-stage review. A short look at a mature design to find late, low-disruption improvements. Suited to subjects with low remaining influence but large cost, where a full study would be wasted.

Budget follows type. A full study on a large building might cost 60,000 to 90,000 CU in team time, facilitation and independent cost checking. A focused workshop might cost a third of that. When the budget cannot cover every candidate, the practitioner asks two questions: which sequence gives the most value, and can some candidates be served by a lighter study type, or combined with another study because they share interfaces?

That last question matters. Two projects that share a plant, a data platform or a site boundary may be better studied together, or at least with a joint session, than separately. We return to portfolio interfaces in Session 3.

Judgment and the limits of scoring

Weighted screens are popular because they make reasoning visible. They also carry well-known weaknesses, and a practitioner should name them openly.

• Scores are ordinal judgments presented as numbers. A 4 is not precisely twice a 2. Small differences in total score, say 74 against 76, are not meaningful; treat candidates within about five points of each other as a tie and decide on other grounds.
• Criteria overlap. Cost magnitude and strategic importance often move together, which can double-count size. Check whether a small but strategic candidate is being buried.
• Optimism and advocacy. Project managers who want a study, or want to avoid one, will argue scores up or down. Independent scoring and written reasons reduce this.
• Missing information. Early-stage candidates often lack estimates. Score them with a stated assumption and flag the confidence, rather than defaulting to a low score.

The ranking is the start of a conversation with the portfolio owner, not the end. Sometimes judgment should override it. A candidate with a lower score may have a decision gate in six weeks that will lock in a costly choice; a study now is worth more than a study later on a higher-scoring project with a longer window. A candidate may also be excluded for reasons the screen does not see, such as a pending policy decision that could cancel it.

Record any override and its reason. Over time, the record of selections, overrides and results lets the value office calibrate its weights against what studies actually delivered.

Running selection as an annual cycle

Selection works best as a regular cycle tied to the portfolio's own planning calendar, not as a one-off scramble when a budget appears. A simple cycle has four steps.

1. Gather candidates. Ask portfolio, programme and operations owners for every project or process that will pass a major decision in the next 12 to 18 months. Include running-cost subjects such as service contracts and processes, not only capital projects.
2. Screen. Score each candidate with written reasons, agree the ranking with a portfolio representative, and flag ties and interfaces.
3. Commit. Agree the study type, timing, sponsor and budget for each selected subject, and tell the unselected project teams why, with any lighter support on offer.
4. Review. At year end, compare each study's implemented results with its screening score. Did high-scoring subjects deliver more? Were any overrides vindicated? Adjust the weights or criteria if the pattern is clear.

The review step is the one most often skipped, and it is what turns the screen from an opinion into evidence. Even three or four years of records will show whether, for example, remaining influence predicts results better than size, which is a common finding.