Vacancy Turn Duration: A Stage-Based Measurement Design

Vacancy cost is usually discussed as a single number of days. That number is useful for forecasting and almost useless for improvement. A twenty-day turn caused by an approval delay needs a different response from a twenty-day turn caused by a back-ordered part. Aggregating them hides the only information that would change the process.

This note proposes a stage-based measurement design for turns. It focuses on definitions, data capture, and interpretation. It does not propose a target duration, because unit condition, market, scope, and season vary too widely for a universal benchmark to be meaningful.

Start With an Explicit Turn Definition

A turn begins at a defined event. Common choices are the vacate date, the key return, or the date the unit becomes legally available. These are not the same. A resident may vacate on the first but return keys on the third, and the operation may not gain access until the fifth.

Pick one definition and record the others. The most useful primary definition is usually the date the operation can begin work, because that is the event the operation controls. Report the gap from vacate to access separately as a distinct measure.

A turn ends when the unit is rent-ready, not when it is rented. Rent-ready should have its own evidence standard: cleaning complete, repairs verified, safety checks done, and the unit presented for listing. Using the lease date as the end point mixes operations with leasing performance.

Split the Turn Into Stages

A practical stage set:

1. Access and condition capture
2. Scope definition
3. Approval
4. Materials and vendor scheduling
5. Work in progress
6. Inspection and correction
7. Cleaning and final preparation
8. Listing ready

Each stage needs an entry condition, an exit condition, and an owner. Without exit conditions, stages become labels. "Scope defined" is not true because a contractor walked through. It is true when the scope, the basis, and the approval path are recorded.

Measure Waiting Separately From Working

Within each stage, record active work and waiting. A unit can sit in "work in progress" for two weeks while the only activity is a vendor waiting for a part. That is waiting, not progress.

Use a small set of wait reasons: access, approval, parts, vendor availability, inspection, cleaning, documentation, and resident coordination. These reasons point to different remedies. A high approval wait suggests delegated authority is too narrow. A high parts wait suggests procurement or specification should change.

Capture the Date That Moves

Target dates change. When a move-in date moves, record the old date, the new date, the reason, and the owner. A date that moves without a reason is the clearest sign that a blocker is being hidden.

Track the number of moves per turn and the total days added. This is often more informative than the original forecast accuracy, because it shows how the operation responds when the plan changes.

Compare Like With Like

Segment turns by scope: cosmetic refresh, standard turnover, and major repair. A portfolio that mixes a two-day paint turn with a six-week rehabilitation will produce an average that describes neither.

Also segment by property type and age. Older buildings generate more surprise conditions, and that is a planning input rather than a performance verdict. Where possible, record surprise discoveries as a separate category so planned and unplanned work can be read apart.

Use the Distribution, Not Just the Mean

Report the median and the tail. The median describes the typical turn. The slowest decile describes the turns that consume the most vacancy cost and generate the most leasing disruption.

A useful view is a stage-level distribution: for each stage, the median and the longest waits. This shows where the outliers originate. Without it, a team may focus on the stage with the highest average while the real tail comes from a single approval bottleneck.

Audit the Evidence

Once a month, sample closed turns and verify that each stage's exit condition is supported. Check that the condition record exists before scope approval, that the scope matches the invoice, that the inspection record exists before cleaning is marked complete, and that the listing date follows rent-ready rather than preceding it.

This audit improves data quality and process discipline at the same time. A stage marked complete without evidence is a data defect and an operating risk.

Communicate With the Same Definitions

Leasing, operations, and owners should see the same stage definitions and the same dates. When leasing sees "ready" while operations sees "awaiting inspection", the resulting promise damages trust. A shared weekly view with confirmed dates only, plus a separate internal view with blockers, prevents most of these conflicts.

What the Measure Can and Cannot Tell You

Stage-based measurement can tell you where time goes and which waits are structural. It cannot tell you whether the market rent is correct, whether the scope was wise, or whether a vendor is fairly priced. Those require separate review.

It also cannot remove the need for judgment about habitable conditions and safety. Measurement should never delay a safety decision. If a unit requires immediate mitigation, the stage clock records the exception rather than blocking the work.

Sources and Method Notes

The stage model and the wait taxonomy are the author's proposed method. For housing condition and household data that can support portfolio baselines, see the American Housing Survey and HUD User. For rental market context, see National Association of Realtors research and Harvard Joint Center for Housing Studies. These sources inform context, not the specific stage definitions above.