Ask an operations team at a 10,000 unit portfolio how long a turn takes and you will get an answer within seconds. Ask what share of that time was driven by the departing resident rather than by a vendor, and the conversation stops.
The turn scorecard in most portfolios is a vendor scorecard. It tracks paint, flooring, make-ready, punch, and the days each trade took. All of that is worth tracking. It also starts counting on the day the unit is empty and available, which means the entire period between notice and an empty unit sits outside the measurement.
That period is not short, and nothing in it is automatic.
What The Industry Actually Measures
The published numbers describe the outcome rather than the cause.
RentCafe, analyzing Yardi data across 139 markets, reported that “apartments take 42 days to fill, on average (up from 40 one year ago)” for market-rate properties of 50 units or more in the second quarter of 2026, with occupancy at 92.9%. On stabilized properties specifically, Yardi Matrix data reported through Multifamily Dive put “the national occupancy rate for stabilized properties” at “94.3% in October, flat MOM and down 0.2% YOY.” The two figures describe different universes and should not be blended.
The cost side is better documented. The National Apartment Association reported that “vacancy and rent loss rose again to $1,323 per unit, up 2.6% from 2023 and marking the fourth consecutive annual increase,” and that leasing expenses reached $292 per unit “largely driven by a 17.5% increase in turnover costs year-over-year.” That analysis covers “over 1 million multifamily units across more than 4,600 properties in 109 metropolitan markets.”
On the turn itself, a survey of 306 multifamily professionals conducted in early 2025 by turn-services provider Rent Ready found that “only 24% turn a unit in 5 days or less” while “51% say that is the ideal,” and that “the largest group, 42%, reports 6 to 10 days.” Their market benchmark work puts the average at “13.1 days to turn a unit.”
Every one of those numbers starts after the resident has gone.
There Is No Benchmark For The Other Half
This is worth stating plainly, because it is the finding. No published industry benchmark exists for the resident-driven portion of the turn. Not from the National Apartment Association, not from the National Multifamily Housing Council, not from Yardi, not from RealPage. The trade press covers vendor wait times in detail and covers the resident side not at all.
An unmeasured process is simply a process nobody has looked at yet.
What Is Actually Inside The Gap
At portfolio scale in residential real estate, the resident-driven half of the turn contains a consistent set of items.
Notice handling comes first. The notice arrives through a portal, an email, or a conversation at the leasing desk. How quickly it is acknowledged, confirmed, and converted into a scheduled move-out date varies by community and by individual.
Move-out scheduling follows. The confirmed date, the time window, and whether the resident holds to it. Late and partial move-outs push every downstream trade.
Physical access coordination sits alongside it. Elevator reservations, loading dock windows, and certificate of insurance collection from the resident’s chosen mover. In dense urban assets this is frequently the binding constraint on the whole schedule.
Key and fob return closes it. A unit with an unreturned key is not an empty unit, whatever the system says.
Left-behind property extends it. Furniture and belongings that were never moved become a disposal task before make-ready can begin.
None of those are maintenance problems. All of them land on the make-ready clock.
Why It Stays Invisible
Three structural reasons, and none of them are about effort.
The first is that there is no system of record. The notice creates a task. The move-out date lives in someone’s calendar or in a note. The dock booking lives in a building email thread. Nothing produces a timestamp that rolls up to a portfolio report.
The second is that the work is human. In Zego’s 2026 Resident Experience Management Report, residents were asked which tasks they expect to handle through a community manager rather than online. For “coordinating move-in or move-out logistics,” 24% said only a human and a further 24% said mainly a human, with just 11% saying mainly digital and 6% only digital. Zego’s own commentary reads: “Moving often comes with high-stress moments with multiple variables. In these moments, residents don’t want a chatbot, they want someone who can be personally accountable.”
Work that depends on a person is work that absorbs staff time quietly. It does not generate a ticket, so it does not generate a number.
The third is ownership. The departing resident belongs to nobody. Leasing has moved on to the next applicant. Maintenance owns the unit once it is empty. Accounting owns the deposit. The window in between has no owner, so it has no target.
The Commercial Consequence
Days are the obvious cost, and NAA’s vacancy and rent loss figure prices them. The less obvious cost is the revenue that moves through the same window.
The departing household is buying movers, packing, and storage on exactly the schedule the operator is trying to coordinate. Insurance sits alongside those services as financial risk mitigation, both for the departing resident at the next address and for the operator receiving a new resident who needs verified coverage before keys change hands. When move-out scheduling is a phone call and a note, those services are booked elsewhere and the coordination value is lost along with the revenue.
Coordinating the move and participating in it are the same operational motion. Running one without the other is what produces an unmeasured window that costs days and earns nothing.
Traditional tools focus on task tracking and administrative coordination. Moved embeds revenue-generating services and insurance verification directly into the workflow.
Four Numbers Worth Instrumenting First
None of these require a new system. They require timestamps on things that already happen.
Notice to confirmed move-out date. The elapsed time between notice arriving and a date being confirmed with the resident in writing. This is the cleanest early indicator of whether the process is automated or remembered.
Scheduled date versus actual vacate date. The spread, tracked as a distribution rather than an average, because the tail is where the damage is.
Access coordination completion. Whether elevator, dock, and mover insurance were confirmed before the move date rather than on it.
Key return to make-ready start. The dead time after the resident has left and before the first trade arrives. In most portfolios nobody currently owns this number.
Hold all four at the portfolio level and compare the spread between the best region and the worst. A wide spread is a process problem rather than a market problem.
Where To Start
Take one quarter of move-outs across the portfolio and reconstruct the timeline for each: notice date, confirmed date, actual vacate, key return, make-ready start. Most operators find they cannot populate the middle columns for the majority of records.
That gap is the finding. From there the sequence is to trigger the move-out workflow automatically off the notice event, capture each step as a timestamp rather than a conversation, standardize it across every region, and report it alongside the vendor-side turn metrics that already exist.
Frequently Asked Questions
Is this just a maintenance scheduling problem?
Maintenance scheduling is downstream of it. The trades cannot start until the unit is empty, the keys are back, and any left-behind property is gone. Those are resident-side events, and they are the ones with no owner.
Our turn times look fine. Why look at this at all?
Turn time as most portfolios define it starts when the unit is available. A portfolio can report a strong make-ready number while losing a week upstream of it. The two measurements do not overlap, so a good one tells you nothing about the other.
Residents are unpredictable. How much of this is actually controllable?
Resident behavior varies. The process around it does not have to. Confirming a date in writing, booking the dock in advance, and collecting mover insurance ahead of the day are all operator-controlled steps, and they are where most of the variance comes from.
What does good look like at 10,000 units and above?
Every notice triggers the same sequence automatically, every step produces a timestamp, and the resulting numbers are reported regionally. The test is whether the process survives a regional manager changing roles.
Closing
The vendor half of the turn is measured, benchmarked, and argued about at every operations meeting. The resident half has no benchmark anywhere in the industry, which means every operator is running it on assumption. The days are real, the cost of them is published, and the first operator to instrument the window will be the first to know what it was actually worth.
If you manage 10,000 units or more and want to see what your own notice-to-key-return window looks like, reach out to our team.
Related reading: how operators turn the move-in and move-out process into a revenue workflow, and the complete guide to resident onboarding automation. For the resident view of the same workflow, see Moved for residents, and for the operator view, see Moved for multifamily.