A healthcare executive scorecard that earns its place in the meeting.

Start with the decisions, not the available fields

An executive scorecard has a limited job: help a leadership team understand the business well enough to decide where attention is needed. It does not need to show every measure available from every connected system.

Start with the recurring operating conversation. Ask the team to list the decisions it expects to make from the report. Then identify the questions that must be answered to support those decisions.

“Understand collections by location” is more useful than “add a revenue dashboard.” It identifies the measure and the level of detail. “Understand why collections changed” adds a different requirement: comparisons, context, and a path to further investigation.

Keep the first scope narrow enough that each number can be defined and checked. A report with six understandable measures can be more useful than a report with thirty ambiguous ones.

Give every metric a contract

A metric contract does not need to be a long technical document. It needs to state the conditions under which the number means what its label says.

Use the following worksheet for every measure you expect leadership to rely on.

The contract makes discussion more productive. If two teams disagree about a number, they can compare definitions rather than debate which screenshot looks right.

Separate financial questions that sound similar

Financial labels often carry more meaning than a source extract can support.

Cash collected, net collections, recognized revenue, new-customer revenue, and lifetime value are different measures. The date, refunds, source basis, and business logic can differ for each.

For an illustrative collections measure, define the qualifying cash activity, refund treatment, location ownership, and business date. Then label it accordingly. Do not rename the result “revenue” simply because that is the label the meeting usually uses.

If the organization needs recognized revenue or margin, establish the additional accounting or cost inputs and the approval owner. If it needs lifetime value, establish the population, horizon, maturity, and value definition.

The useful practice is to preserve the distinction. A narrow, well-defined measure can support a good decision even when a more ambitious measure is not ready.

Keep people and ratios from being added incorrectly

Some portfolio measures are additive; others are not.

Consider a fictional person who pays at two locations during September. They can legitimately appear in each location’s payer count and still represent one distinct payer in the network.

Adding the two local counts would answer a different question. The network distinct count needs to be calculated for the network population.

The same care applies to rates. Suppose one fictional location has 10 qualifying outcomes from 20 eligible events and another has 10 from 100. Averaging their percentages produces 30%, while combining the populations produces 20 outcomes from 120 events, or about 16.7%. Which result is useful depends on the intended question; they are not interchangeable.

Document how a metric behaves when a user changes the scope. Test it with an example that exposes the difference, rather than only checking a happy-path total.

Build a coherent first screen

A useful first screen often needs three kinds of information:

1. Scale: the selected period’s defined financial or operating amount.
2. Change: a comparison with an appropriate prior period or target.
3. Coverage: what is included, complete, or unavailable.

Then provide a route into the detail. That might be a region, location, service, source, or cohort view, depending on the measure.

The supporting detail should remain consistent with the selection. If a companion chart uses a different period, say so. If the region filter does not apply to an advertising dataset, do not let the interface imply that it does.

Design the screen around a reading order. What should the reader notice first? What context prevents a wrong conclusion? Which action or next question should follow?

Compare periods deliberately

Period comparisons can change meaning when the underlying scope changes.

A partial month compared with a complete month is not the same as two complete months. A network with newly opened locations is not automatically a like-for-like comparison. A source with a delayed refresh may have fewer included business days than another.

State the basis of the comparison. Identify when the measure uses calendar months, equal-length windows, a cohort period, or another convention.

When the business wants several perspectives, give them separate labels. A total-network change and a same-location change can both be useful, provided the report tells the reader which one is shown.

Avoid silently changing the comparison rule to make a chart look cleaner.

Show missing information as information

An unavailable value is not a zero. A delayed source is not necessarily a weak location. A withheld small group is not an empty group.

The scorecard should have readable states for these conditions, with a short explanation of what the reader can and cannot infer.

For example:

• “Unavailable: the source does not provide this measure for the selected scope.”
• “Partial period: publication is complete through the stated date.”
• “Withheld: the approved reporting rule does not permit this result.”
• “Definition pending: the business owner has not approved the measure.”

Each state should point to an operational responsibility. Who investigates a delayed source? Who approves a mapping? Who decides whether a requested metric belongs in the scope?

Keep marketing delivery and acquisition outcomes distinct

Ad spend, clicks, CRM records, appointments, and purchases can appear in the same operating discussion. They may still represent different populations and observation windows.

Do not construct a funnel merely because the labels can be arranged left to right. Establish the qualifying entry population, event definitions, identity mapping, exclusions, and time windows.

Until the data supports that relationship, keep the source measures distinct. A delivery view can still help the growth team ask useful questions without pretending to prove a customer outcome.

When a derived measure becomes ready, include its contract and validation in the reporting change rather than silently replacing an older label.

Use a repeatable review sequence

Try a short operating-review routine:

1. Confirm the period, scope, and coverage.
2. Identify the largest meaningful changes.
3. Check whether a definition or source issue could explain them.
4. Open the relevant detail.
5. Separate observed facts from possible explanations.
6. Assign the investigation or action to a person.
7. Record any reporting change required.

The scorecard should support this sequence without claiming to make the decision itself. Its job is to make the question clearer and the evidence easier to inspect.

A readiness checklist

Before adding a metric to the executive review, check that it has a defined purpose, an approved population, a clear period, an understandable calculation, a source owner, a meaningful comparison, and a visible limitation.

Before replacing an existing report, reconcile the result and explain any differences. A corrected definition may intentionally produce a different number; the team needs to understand why.

Finally, name the person responsible for keeping the scorecard useful. The first version is only the start of the reporting routine.

Explore the fictional reporting example to see how period, scope, definitions, and unavailable measures can sit beside the chart.