A list of efficiency measures is not yet an investable package

## Ask what the total assumes about the parts

I approved the total because the underlying work looked serious. Twelve standalone efficiency estimates had been assembled into a portfolio paper, and the sum was 19.2 GWh a year. Each estimate had a purpose. What I failed to ask was whether the measures could be combined on the basis the total implied. Several affected the same system. Their individual accounts did not establish the benefit of implementing them together. The arithmetic was not the difficult part; the strategic interpretation was.

An efficiency portfolio can contain changes that alter one another's conditions. A measure examined on its own may face a different setting after another measure is introduced. Qualified specialists need to examine those relationships and the assumptions of the proposed package. A strategist should not attempt to replace that assessment with a spreadsheet shortcut. My responsibility is to recognise when the business is relying on a combined proposition and to require an account appropriate to that proposition before presenting its apparent scale as a basis for commitment.

There is a reasonable argument for using standalone estimates early. They can identify areas worth exploring, help compare possible directions and make a large field of opportunities understandable. I would not discard them. The important question is what status they carry in the decision. A list for investigation is different from a package proposed for delivery. As the proposal moves towards a commitment, the account needs to change accordingly. Otherwise a useful early screening device becomes a stronger claim without anyone explicitly accepting responsibility for the change.

That transition is easy to miss because the same table can survive several meetings. A row first described as a possibility becomes part of a plan; the total then appears in a funding paper or customer conversation. The qualifiers may remain in a supporting file while the headline travels alone. I now ask what the reader is being invited to conclude from the total at each stage. If the answer has changed from exploration to a deliverable proposition, the supporting evidence must be examined for that new purpose.

In our fictional case, Hamid's technical challenge made the gap explicit. I could have described it as a late technical refinement and left the portfolio story largely untouched. That would have understated my role in the problem. I had approved the larger interpretation and needed to withdraw it. A strategy function is accountable for how it assembles the account, even when the individual technical estimates were prepared by others. Combining evidence is itself a consequential act when the combination changes what the business believes it can do.

## Keep the revised forecast in its proper category

The review cost ₹72 lakh and produced a package model of 11.4 GWh a year under stated assumptions. That figure was more appropriate to the proposed combination, but it was still a model. It was not an observed reduction in customer consumption, a verified financial saving or an emissions result. Correcting one interpretation should not introduce another by allowing the revised number to sound as though the project has already delivered. The categories need to remain clear even when management is relieved to have a more defensible estimate.

The difference between 19.2 and 11.4 also required care. It was not energy physically lost during the review, nor a quantified benefit produced by the strategy team. The first figure was an unsupported aggregation for the combined purpose; the second described the examined package model. Treating the gap as a realised outcome would have created a new story from a correction to the old one. I wanted the directors to understand why the account had changed and what decision the new account could support.

Forecasts remain useful despite their uncertainty. The alternative is not to wait for every operating result before making any investment. It is to explain the assumptions, the relevant technical assessment and the stage of commitment being considered. A model can support a pilot or a further investigation without proving the full proposition. Management should be able to say what it expects to learn and how different findings would affect the next choice. That makes uncertainty part of the strategy rather than a generic disclaimer attached to its end.

The cost of reaching that position should also remain visible. The ₹72 lakh review consumed resources and delayed a more expansive presentation. Some colleagues worried that a smaller forecast would reduce interest in the opportunity. They were right that the revised account was less immediately impressive. I did not try to recover the original scale through a more ambitious financial interpretation. The company needed to decide whether the narrower proposition was worth pursuing, including the cost already incurred and the possibility that further work might narrow it again.

It matters who owns the related conclusions. Engineering specialists assess the interaction of measures and the measurement approach. Finance examines the financial case and its treatment. Commercial owners establish the actual customer proposition and terms. Strategy connects these accounts into a choice for the authorised decision-makers. The function becomes less useful when it either treats every specialist finding as an unquestionable finished answer or assumes that it can resolve every disagreement itself. The aim is a coherent decision with visible responsibility, not a single department claiming all the expertise.

## Choose the next commitment around the package

We funded three bounded pilots with ₹2.6 crore approved expenditure. That was the next commitment, not a statement that the full portfolio had been delivered. The pilot scope needed to reflect the package questions we wanted examined. A pilot chosen merely because it was easy to present might leave the important interactions unanswered. I asked the responsible owners to explain how the work would inform a later decision, while preserving their authority over design, operation and evaluation. The strategic account should identify the purpose of learning without prescribing the engineering.

The narrower route surrendered some speed and scale. A larger initial programme might have secured attention or resources that a smaller one did not. Those were genuine trade-offs, and a credible recommendation should acknowledge them. I did not need to portray every supporter of the original plan as careless. The useful comparison was between a commitment supported by the examined package and a broader commitment still relying on unresolved assumptions. Management could then consider the opportunity and exposure in each route, rather than choose between an attractive plan and an abstract warning.

Directors can ask whether the portfolio total describes independent opportunities, interacting measures or a package already assessed as a combination. They can ask which assumptions change when the scope changes, what the next expenditure is intended to establish and how a contrary result would be reported. These are oversight questions. They do not require directors to calculate an energy model or supervise installation. They require management to explain the business meaning of the evidence in a form that supports collective judgement.

The experience changed what I look for in a strategy paper. A long list can demonstrate effort without establishing a coherent enterprise choice. The connection between the parts is often where the important question sits. In efficiency work, that may concern the interaction of measures; in another business, it may concern shared customers, resources or obligations. The principle is to examine the proposed combination on its own terms. A total should help the company understand the commitment it is considering, rather than borrow the credibility of each individual row and assume that adding them has completed the strategy.