A league table cannot repair an inconsistent denominator

A ranking gives a management discussion an immediate structure. Sites appear in order, a gap is visible and somebody can be asked to explain why they are near the bottom. That clarity is useful only if the comparison has a defensible basis. When the denominator means different things across the sites, the table can turn a definition problem into an apparent performance problem before anyone has examined the underlying activity.

At Relzavi Clinical Linen Services, a proposed consumption comparison covered eleven contract sites. Four used occupied-bed-day definitions that were inconsistent with the others. The ranking therefore could not support the efficiency conclusion being proposed without further work. A precise ratio did not resolve the fact that its denominator had been constructed differently in part of the population.

The contract team wanted to publish the table with a note explaining variation so that the performance conversation could begin. I argued that the difference affected the central comparison rather than a minor qualification. We postponed the ranking and funded ₹24 lakh of definition work with contract and operating owners. That choice delayed a visible output and required agreement among people whose existing collection practices served different purposes.

## Ask what the denominator represents

A denominator is not merely the lower half of a calculation. It defines the activity against which the numerator is being interpreted. In our case, occupied bed days were intended to provide a basis for discussing linen consumption. Before using the ratio, we needed to understand how each site counted that activity and whether the definitions were sufficiently aligned for the proposed comparison. The arithmetic could be correct while the interpretation remained unsound.

There may be sensible reasons for different local definitions. A site might collect activity for a contract, an operating plan or another reporting obligation. The aim of the review is not to declare every existing practice wrong. It is to establish whether those practices can support this particular comparison and, if not, what agreed basis is needed. A data leader should be able to explain that purpose without asking every team to abandon useful local records unnecessarily.

The work also needs owners who understand the operations behind the count. A central data team can identify inconsistencies and propose a definition, but the people collecting and using the activity must examine whether it is workable and accurately described. Otherwise, agreement may exist only in a document while daily practice continues differently. The definition should survive the ordinary work of recording activity, not merely the meeting in which it was approved.

The contract team's concern about delay was real. Managers wanted a way to discuss differences and identify areas for attention. Postponing the ranking did not mean that all performance discussion had to stop. We could still describe observed activity and the uncertainty in the comparison. What we withheld was the ordered judgement that the inconsistent denominator could not yet support. That distinction allowed work to continue without giving the table more authority than the evidence justified.

A qualification is useful when it helps the reader interpret a result that remains meaningful. It is less useful when it asks the reader to mentally undo the main conclusion. If the first impression is a clear ranking and the note says the underlying activity is counted differently, the document may still encourage an action the evidence cannot sustain. I wanted the form of the report to match the strength of the conclusion, not rely on careful readers to repair it themselves.

The cost of definition work should also be visible. The ₹24 lakh programme was not an automatic efficiency saving. It paid for a more consistent basis on which later decisions could be made. That can be a worthwhile investment, but it should be assessed as such. Inflating the benefit by claiming a consumption reduction before one has been established would repeat the same error we were trying to avoid: allowing a useful data improvement to stand for a broader operating outcome.

## An agreed comparison is a beginning

All eleven sites subsequently used an agreed basis for the comparison. That addressed the identified denominator inconsistency. It did not prove that every remaining difference reflected efficiency. Service mix, operating conditions and the way linen was used could still matter. A better measure made the next question more meaningful; it did not remove the need for interpretation or supply the answer on its own.

The distinction matters when management turns a comparison into a performance intervention. A site should not be asked to reduce a ratio without understanding what change in work would produce that result and what other consequences might follow. Directors can ask whether the proposed action is supported by the analysis, whether relevant operating differences have been examined and who is responsible for evaluating the result. A ranking can identify a question; it should not automatically prescribe the remedy.

I also pay attention to how the new definition is maintained. A comparison may be valid at one point and drift as collection practices change. A new contract, a revised operating process or a different local reporting need can alter the meaning again. The organisation should know who can approve changes and how their effect on the comparison will be reviewed. Consistency is a continuing responsibility, not a certificate issued at the end of one programme.

For a board, this is an example of why simple measures sometimes require substantial judgement behind them. Directors should not have to reconstruct every denominator, but they can expect management to explain the basis of a consequential comparison and the limits of its conclusion. A report becomes more useful when it shows what is comparable, what remains uncertain and what further evidence is needed before a decision is made.

The final value of our work was a more honest performance conversation. We gave up the immediate convenience of a league table and established an agreed basis across eleven sites. The resulting measure could support discussion without pretending that every difference had already been explained. I would rather bring a board a comparison with a clear boundary than an impressive ranking whose order depends partly on definitions the business has not reconciled. Precision in the ratio is useful only after precision in its meaning.