Property management KPIs: build a report people use

A property management KPI dashboard should help someone decide where to spend attention. Choose the decisions first: which arrival needs intervention, which cost needs investigation, which future dates need a revenue review. Then choose the smallest set of measures that makes those decisions easier.

Occupancy, revenue and completed work can all look reassuring while important problems remain open. The dashboard needs the definitions and exceptions behind the headline values. AI can summarize changes after those numbers are calculated, but it should not invent the definitions or the reason for a movement.

This guide focuses on rental and hospitality operating reports. Financial measures and their accounting treatment should be approved by the responsible accountant. Use the reporting automation guide for the wider source and release process.

Start with a decision-to-measure map

Ask each reader what action they could take after seeing a value. If nobody can explain the action, the measure may belong in a reference report rather than the main dashboard.

Don't force a daily operations board and a monthly financial report into the same freshness promise. A manager may need a current job queue and approved prior-month expenses on one screen. Label the periods separately so the layout doesn't imply the figures were all updated together.

The daily operations report guide gives more detail on the short exception list that should sit near the top.

Define the unit of analysis

A listing, physical unit, reservation and property are not necessarily the same thing. Decide which one the metric measures. Otherwise, a new listing version or combined unit can change the denominator without adding capacity.

Keep one row's meaning consistent in each calculation. Microsoft's data-modeling guidance recommends consistent detail within fact tables. In practice, count work orders from work-order records and booked nights from the appropriate reservation-night records before combining their summaries.

For every KPI, record its numerator, denominator, period, exclusions and source. Include the treatment of inactive properties, owner stays, canceled reservations and blocked inventory where relevant. The operating and finance owners should approve these choices.

The label should tell the reader when a measure differs from a familiar benchmark. A net-payout-based nightly measure should not quietly inherit the meaning of a room-revenue-based industry ratio.

Calculate occupancy and rates from the underlying totals

The CoStar STR glossary defines occupancy using rooms sold and rooms available, ADR using room revenue and rooms sold, and RevPAR using room revenue and available rooms. For a vacation-rental dashboard, document the corresponding unit-night and revenue definitions before adapting these ratios.

The following fictional example uses two groups of rental units over one 30-day period. All units are available throughout the period. Revenue means the same approved lodging-revenue measure in both rows; the example excludes ancillary fees and taxes.

A simple average of 60% and 40% is 50%. That gives the one-unit group the same weight as the nine-unit group. Combined occupancy is 126 divided by 300, or 42%.

For this example, the combined average rate per sold unit-night is $13,500 divided by 126, or about $107.14. Revenue per available unit-night is $13,500 divided by 300, or $45. Calculate with the underlying values and round the final display, rather than multiplying already-rounded ratios.

The same issue affects many other dashboard ratios. A property with one completed task should not necessarily receive equal weight to a property with fifty tasks when calculating a portfolio-wide completion measure. State whether the report shows an average property result or an overall activity result.

Make inventory changes visible

Blocked nights can change a ratio without creating another booking. If your approved definition excludes a category of unavailable nights, report those nights alongside occupancy so readers can understand the movement.

Keep a separate record for why inventory became unavailable. Maintenance, owner use and a temporary distribution problem call for different actions. A single blocked-night total may be enough for a headline, but its detail should be accessible.

Also distinguish same-property performance from changes caused by adding or removing units. A growing portfolio can report higher total revenue while comparable properties perform worse. Neither measure is inherently wrong; they answer different questions.

The revenue reporting guide shows how to describe those differences without turning an arithmetic change into an unsupported explanation.

Check whether two dashboards actually use the same rules

Two reports with the same title can disagree because their definitions differ. Check dates, status filters, currency and source freshness before assuming an integration is broken.

Guesty's comparison of its analytics products documents differences in current-day availability, reservation statuses and date handling. That is a useful reminder to verify the specific configuration behind a value rather than relying on its label.

Create a reconciliation note for each headline metric during setup. It should show the source comparison, approved adjustments and remaining differences. Preserve it when definitions change so future reviewers can explain a break in the trend.

Display an extraction timestamp and the business period covered. If a feed fails, keep the stale status visible. Updating the page header to today's date should not conceal that the numbers came from yesterday's export.

Show the work that has not finished

Average completion time usually describes completed cases. It can improve while difficult cases accumulate in the open queue. Pair it with open-case age and the count of overdue work.

For a service dashboard, consider a panel with completed requests by type, reopened requests, overdue open requests and the oldest unresolved item requiring action. The exact definitions depend on your service process, but the display should not reward closing easy tickets while hard ones remain unattended.

Use sample size with percentages. A 100% result from two reviewed cases means something different from the same percentage across a large, varied population. Avoid ranking small groups as though they had equivalent evidence.

Assign a person to each exception category. A red tile that nobody owns becomes decoration. Let the reader open the relevant records and identify the next step, with permissions appropriate to their role.

Review the dashboard using real decisions

Test one operating meeting with the proposed dashboard. Ask the participants to trace a headline number, explain a movement and choose an action. Note where they leave the report to rebuild a spreadsheet or ask for information that should already be present.

Check a renamed property, a canceled reservation, a repeated import and a failed source refresh. Confirm that each case changes the report as the definition says it should. Use the data quality checks to resolve shared identity problems before adding more charts.

Measure preparation time, time spent disputing definitions and the number of actions that reach a clear owner. Review unused measures after several cycles. Remove a tile if it adds work without informing a decision, and keep the drill-down detail that makes the remaining numbers credible.