A common dashboard can preserve an uncommon unit

## Read the quantity before choosing the chart

The reporting labels were familiar enough that readers assumed they described the same thing. In eight of twenty-eight reviewed measures, that assumption did not hold. Some local records expressed linear metres. Others described saleable rolls, with lengths that varied by product and order. A dashboard could display both without a technical error and still encourage an invalid comparison. The calculation was doing what it had been asked to do. The missing work concerned the business quantity that the calculation was supposed to represent.

I find this problem easy to underestimate because it appears small beside a major systems investment. A unit definition may occupy a short line in a specification, while the programme discussion concentrates on architecture, access and delivery dates. Yet that line can determine what an executive believes about the operation. If the reader interprets a change in roll count as a change in metres, the system has helped create a business inference that the underlying records do not support. More frequent updates only reproduce that problem more often.

Our manufacturing context made the difference important. Saleable rolls were not a universal fixed-length equivalent that could be converted by a single factor without considering product and order conditions. Operations colleagues understood that in their daily work. Group reporting colleagues wanted a common view that could be compared across the business. Both needs were legitimate. I wanted the systems proposal to make the relationship explicit rather than let one function's convenient label become the other function's assumed definition. The shared display had to follow an agreed meaning.

We began with the reviewed measures and the decisions their readers intended to make. That kept the discussion from becoming an abstract argument about perfect data. A measure used to understand physical production may require a different account from one used to plan saleable units. The business can need both. The question is whether the report identifies which quantity it presents and what a comparison across sites or periods actually means. I asked the owners to explain that before the team selected a conversion or changed the chart.

## Make the comparison a business decision

The faster proposal was a standard conversion that would allow all eight measures to join a uniform group display. It had practical appeal. Readers would not need to learn another distinction, and the report could preserve its existing layout. Operations resisted because the proposed factor would conceal differences in the underlying products and orders. I supported a definition, mapping and correction programme costing thirty-eight lakh rupees, with finance and operations owning the intended meaning. IT would implement the decision and help test whether the output reflected it.

Six measures became comparable under the approved definitions. Two remained separately labelled. The result was less uniform than the original aspiration, but it was more faithful to the business. We did not claim that the two had been standardised merely because the system could display them beside the others. Nor did we discard the local quantities that people still needed to run the operation. A common reporting environment should be capable of preserving a meaningful distinction rather than treat every difference as an implementation defect to remove.

The expenditure covered mapping, report changes and follow-up validation. It also required business time to settle definitions that could not be delegated entirely to the technical team. That effort is easy to omit from a systems budget because it does not look like software construction. In our case it was central to the work. Without it, the developers would have been asked to encode an unresolved disagreement. The eventual dashboard might have looked finished while its most consequential assumption remained unowned.

Finance retained accounting policy and operations retained responsibility for the production meaning. My role was to keep those authorities visible in the systems decision. A CIO should not settle a disagreement about a business quantity by choosing whichever field is easiest to aggregate. Equally, the technical team should not leave the disagreement hidden until final acceptance. It can expose the consequences of each definition, show where the data supports a proposed comparison and make the unresolved choice concrete enough for the responsible owners to decide.

## Let the remaining difference stay visible

The two separately labelled measures required explanation in the executive view. I regarded that as part of the work, not an embarrassment after an otherwise successful programme. A reader should know why the comparison has a boundary and what would have to change before the measures could be combined. The explanation need not reproduce every operational detail. It should prevent a reasonable reader from inferring a uniformity that the underlying process does not have. A short, accurate qualification can be more useful than a visually seamless total.

There is a temptation to describe six resolved measures as progress toward an inevitable eight. That may be the wrong story. The remaining difference can reflect a real business need that should continue, rather than incomplete implementation. I wanted management to distinguish work still required from variation it had deliberately retained. If a future operating change makes the two measures comparable, the report can change with it. Until then, the system should not anticipate that future state by presenting it as though it already exists.

For a board, the relevant questions concern the decisions made from the report and the ownership of its definitions. Directors do not need to choose field mappings. They can ask whether a consolidated number combines like quantities, what assumptions permit the comparison and which important exceptions remain. They can also examine whether management has funded the business work needed to support the systems promise. A dashboard project that includes no time for definition decisions may be depending on agreement that has not yet been reached.

Our review covered twenty-eight measures, found eight with unit problems and corrected six into comparable definitions while retaining two separately labelled. The thirty-eight-lakh-rupee programme did not certify every group report. It established a more defensible account for the set examined. The experience strengthened my belief that a reporting system should help a reader make the right distinction, even when that distinction complicates the display. Common infrastructure is valuable. Common meaning must still be established by the people who understand the business quantity, and the system should preserve the places where that meaning remains different.