A supplier questionnaire can produce a reassuring view of progress. Fields are filled, the completion rate rises and a previously fragmented subject begins to look manageable. I find that useful as an account of collection work, but incomplete as an account of evidence. A number in a field may have been measured, modelled or derived from an assumption whose scope is not yet understood. The organisation needs to know which before the uniform appearance of the form becomes a uniform level of confidence in the result.
Our fictional 2025 review at Tervanex Components covered 270 suppliers. One hundred and forty-four provided measurements, fifty-nine supplied modelled estimates and sixty-seven had no submission. Those categories reconciled to the population, but they did not settle the quality or comparability of the received values. A measurement supplied by a company was not automatically independently verified. A modelled estimate was not necessarily useless. A missing response established neither a high footprint nor a zero. The data account needed room for all three distinctions.
## Coverage describes receipt before it describes confidence
I would begin by separating the work of obtaining a response from the work of understanding it. Both can require effort, and neither should disappear into the other. Procurement colleagues may have made substantial progress in engaging suppliers, while specialists still need to examine boundaries, periods and supporting methods. A leadership report can acknowledge the collection achievement without implying that every received value is ready to support the same conclusion. The distinction is especially important when the next audience did not take part in the collection process.
The commercial team wanted a consistent account for customer conversations. That was a legitimate need. An organisation cannot respond usefully if every discussion begins with an unexplained pile of unrelated records. The question was what kind of consistency we should pursue. Common presentation could make the state of the evidence easier to understand, or it could conceal differences in order to make the account look finished. I supported the former, even where it required a more qualified explanation than a single complete-looking total.
There are several reasons why two received values may need further examination before being combined. Their described boundaries may differ, their periods may not align or the underlying activity may not be defined in the same way. The CDO need not claim expertise in every footprint method to recognise that these questions matter. My function's responsibility was to preserve the relevant basis and route the question to those qualified to assess it. A consistent database cannot, on its own, make differently defined measurements comparable.
Nor should the word measured carry an assurance conclusion it has not earned. A supplier can provide a measurement that still requires explanation or review. The status records something about the stated source of the value; it does not automatically tell the recipient that every supporting process has been independently examined. That qualification is not an accusation against the supplier. It is a description of what the receiving institution knows, and an invitation to distinguish evidence supplied from conclusions yet to be reached.
## An estimate can be useful without changing its status
The fifty-nine modelled submissions prompted a different discussion. Some colleagues feared that showing them separately would make the account appear weaker. Others wanted to exclude them entirely, which could discard information relevant to a defined purpose. I did not regard either response as automatically right. An estimate can be useful if its basis, assumptions and intended use are understood. The problem arises when its usefulness is defended by quietly presenting it as a measurement or by removing the conditions on which it depends.
Keeping the status visible allows a better conversation about the decision. A preliminary comparison may reasonably use material that would be insufficient for a stronger public statement. The institution needs the appropriate professional judgement about that use, rather than a data-team declaration that the existence of a number settles the matter. If a different purpose requires stronger evidence, the work remains to be done. The earlier estimate can still be described honestly as part of the organisation's learning without being retrospectively promoted into something it was not.
I supported S$4.6 million in data-review work. That expenditure helped organise the submissions and their qualifications for the responsible procurement, sustainability and assurance colleagues. It did not give my function the authority to issue an assurance opinion. I would keep the distinction close to the figure in a leadership profile because otherwise a substantial programme budget can create an impression of broader authority than the executive actually held. The scale of the effort and the nature of the conclusion are separate facts.
The review also needed to avoid an unhelpful hierarchy in which every measured value was treated as superior for every purpose and every modelled value as inferior. The question was fitness for the proposed use, with the relevant boundaries examined. A poorly understood measurement may not answer the intended question merely because it occupies the preferred category. Preserving the categories creates a basis for scrutiny; it does not complete that scrutiny or substitute a simple ranking for specialist assessment of the actual evidence.
## Absence should remain visible in the account presented
The sixty-seven missing submissions were particularly important to preserve. A blank can disappear when data is aggregated, leaving a total that looks lower or more complete than its basis permits. Treating absence as zero would make the missing evidence influence the conclusion in a direction it did not justify. I wanted the account to identify what had not been received and why the resulting coverage remained incomplete. That statement could be made without inventing a value for the missing material or assuming a reason for every non-response.
There was a cost to this clarity in the customer-facing explanation. The institution had to discuss uncertainty instead of presenting a single apparently settled position. I regarded that as part of the work, not a communications defect to remove after the analysis was finished. A qualified account can still be useful. It tells the next reader which questions have been addressed, which remain with professional reviewers and which are waiting for evidence. The alternative may be easier to repeat while being harder to defend.
No emissions-reduction result is claimed in this fictional case. Improving the representation of submissions does not itself establish that the underlying footprint fell. Nor does the account describe a later completion result or a net-zero achievement. Those would require evidence beyond the review recorded here. A leadership narrative should be able to show that a decision improved the integrity of an account without adding a more dramatic outcome simply because the subject attracts ambitious language.
For a CDO, the broader responsibility is to preserve the distinction between what the organisation has collected and what it is entitled to conclude from it. The work often happens in details that a summary is tempted to remove: a source category, an absent submission or a boundary still under examination. I would make those details legible enough for other professionals to use, then respect their authority over the conclusion. A completed field is the beginning of that conversation, not evidence that the conversation has already been resolved.