Open your budget and find the line item for turnover.

You will probably find make-ready. Paint, clean, carpet, the odd appliance. Maybe a few hundred dollars of marketing sitting somewhere else entirely.

Now ask what that unit actually cost you between the old resident handing back keys and the new one paying rent. The answer can be substantially higher, because some of the economic impact never appears as a turnover expense at all.

That distinction is the whole article, and it explains something otherwise puzzling: why an industry that agrees turnover is expensive keeps under-investing in fixing it.

The number everyone quotes, and where it came from

A widely cited figure puts average apartment turnover cost at $3,872.

It is worth knowing what sits behind that before you put it in a business case. It comes from a survey of 630 property managers at communities of 250 units or more, run by a multifamily software provider and reported by Multifamily Dive in October 2023. It covers advertising and marketing, repairs, concessions and the cost of lost rent.

Two things about it.

First, the figure continues to appear in newer articles, sometimes without its 2023 research date being made clear. The same provider's current materials cite approximately $3,976 per resident turnover, so $3,872 is best treated as a historical 2023 figure rather than current benchmark data. We saw the same pattern of a number aging in place while its date gets refreshed with the 86% self-guided tour statistic.

Second, and more usefully: it is the all-in economic cost, not the accounting cost. That is why it is so much higher than what your books show.

It is probably not your number

Here is the part that matters more than the average.

In a National Apartment Association survey, 53.8% of respondents reported average turn costs between $1,500 and $3,500 per unit, while another portion reported costs below $1,500 and a smaller share reported costs above $3,500.

That is a useful reminder that a widely quoted industry figure does not necessarily describe your portfolio. Your number depends on your rent, your turn scope, your market, your concession practice and how long the unit sits.

So the honest use of $3,872 is as a historical reference point to sanity-check against, not a number to plan around. Which means you have to build your own, and the rest of this article is how.

Four places the cost lands, and only one looks like turnover

Here is why the number is hard to see in your own accounts. The cost of a turn lands in four places, and they behave completely differently.

Only the first is normally easy to identify directly as turnover work. The others may be recorded elsewhere, or may not appear as an expense at all.

Lost rent is not an invoice. It is rental revenue that was never collected during the vacancy. Concessions typically reduce the revenue recognised from the lease rather than appearing as a separate turnover expense, and the exact treatment varies by operator. Neither belongs automatically on the same expense line as make-ready work, and neither will show up there no matter how carefully you code it.

The effect of that is predictable. Costs that appear as expenses get scrutinised, budgeted and negotiated. Costs that appear as absent revenue get described as market conditions.

Build your own number

Six inputs. Some you have to hand, some need assembling.

Monthly rent. The actual achieved rent, not the asking rent.

Vacancy days. The period the unit was unavailable for rent, from the departing tenancy ending to the new tenancy's rent commencement, using your own vacancy convention.

Make-ready spend. Your actual average, not a benchmark.

Attributable marketing and leasing cost. Advertising spend plus whatever leasing time you can reasonably attribute.

Concession value. Total dollars given on the new lease, whether as free weeks or reduced rent.

Any other directly attributable cost. Locks, inspections, temporary utilities, if you track them.

Then:

Lost rent = (monthly rent ÷ 30) × vacancy days
Estimated economic cost of a turn = lost rent + make-ready + attributable marketing and leasing + concession value + other attributable costs
Annual exposure = average economic cost per turn × expected turns per year

A worked example, with round numbers so you can follow the shape rather than adopt the figures.

A unit at $1,650 loses $55 a day empty. At 38 vacancy days that is $2,090 in rent that never arrived. Add $1,400 of make-ready, $200 of marketing and a half-month concession worth $825, and the estimated economic cost is $4,515.

In that example the make-ready line represents roughly a third of the calculated total. The proportion will vary considerably by property, unit condition, rent level, vacancy period and concession strategy, which is exactly why the exercise is worth doing on your own numbers rather than borrowing anyone else's.

Scale it. If a 200-unit portfolio experiences 86 turns in a year at that cost, annual exposure is roughly $388,000, of which around $120,000 appears in the maintenance budget and the remainder is not represented by that line.

What a day actually costs

The daily figure is worth isolating, because it makes speed arguments concrete.

At $1,650 a month, a vacant day costs $55. Ten days across 86 turns is roughly $47,000, on a portfolio where nothing else changed.

That is what makes the days-on-market split worth producing. If ten of those 38 vacancy days occur before the unit is ready to market, you can quantify how much vacancy-related revenue was associated with the make-ready period rather than the post-ready leasing period. Those two periods have different owners and different causes, and the blended figure cannot separate them. The measurement argument is in our piece on days on market.

Why you probably cannot produce this today

Everything above requires several figures attributed to a single unit and a single turn. Rent, vacancy days, make-ready spend, marketing, concession value.

In most operations those live in different places. Rent in the lease record. Vacancy days reconstructed from move-out and move-in dates. Make-ready spend in work orders, possibly split across several. Concessions in the lease terms, and often not totalled anywhere.

So producing a per-turn cost means someone reconciling several systems, per unit, which is a project rather than a report. It gets done once for a board pack and never becomes a monthly number. Same structural problem we covered in why your response time number is probably wrong.

And the consequence is specific. When you cannot produce your own number, you quote somebody else's. Which is how a 2023 survey figure ends up in a 2026 business case for a portfolio it never measured.

What to instrument

• Economic cost per turn, by property. The number this article exists to produce.

• The visible share. What proportion of your calculated cost appears in maintenance and turnover expense lines. Whatever that proportion turns out to be, it tells you how much of the problem your budget conversations have been covering.

• Lost rent as a separate line. Not an expense, but it belongs on the same page as the expenses so the comparison is available.

• Concession value per turn. Frequently the least-tracked input and often larger than the marketing spend.

• Cost per turn alongside turn volume. A portfolio with a low cost per turn and a high number of turns can be losing more than one with the reverse.

Common mistakes

• Budgeting turnover as a maintenance line. It leaves a significant share of the economic impact invisible to the people deciding how much to spend on preventing it.

• Adopting $3,872 as your number. In NAA survey data most respondents reported between $1,500 and $3,500 per unit. A widely quoted average is not a description of your portfolio.

• Quoting the figure as current data. Its methodology is from a 2023 survey. Somebody in the room may check.

• Ignoring concessions. They do not feel like a cost because no money leaves the building. They are a direct reduction in the revenue the turn was supposed to produce.

• Treating all vacancy days as leasing days. Until the pre-ready portion is separated out, you cannot say which part of the operation the cost sits with.

Back to that budget line. It was never wrong. It was answering the question your accounting system is built to answer, which is what did we spend, not what did this cost us.

The gap between those two questions is where a large part of the money in a turn lives, and it stays invisible for a structural reason: the figures that would close it sit in different places, and nobody has the afternoon to reconcile them for every unit.

That is one of the operational problems RIOO is designed to address. RIOO brings leasing, unit, maintenance, move-in and move-out and financial information into the same property management environment, with dashboards and reports built around that operational data. That gives operators a foundation for building property-specific turnover reporting rather than relying on an industry average.

Run the calculation on your last ten turns and compare the total against what appears in your maintenance and turnover expense lines. The gap will tell you more about your turnover economics than any published figure.

Frequently asked questions

Q1. How much does tenant turnover cost per unit?
A widely cited figure is $3,872, from a 2023 survey of 630 property managers at communities of 250+ units, covering marketing, repairs, concessions and lost rent. NAA survey data has most respondents reporting between $1,500 and $3,500 per unit, so the quoted average does not describe every portfolio.

Q2. Why is turnover more expensive than my budget shows?
Because make-ready and some marketing appear as expenses while lost rent and concessions generally do not. Lost rent is revenue never collected, and concessions usually reduce lease revenue rather than adding a separate cost line.

Q3. How do I calculate the cost of a vacant unit?
Lost rent is monthly rent divided by 30, multiplied by vacancy days. Add make-ready spend, attributable marketing and leasing cost, the value of any concession on the new lease, and any other directly attributable costs. Multiply the average by expected annual turns for portfolio exposure.

Q4. What is a normal turnover rate?
There is no single normal rate that applies across property types and markets. Turnover varies by asset type, resident profile, market, lease structure and operating conditions, which is why comparing your own trend over time is more useful than comparing against a national figure.

Q5. Should I use the industry average in my business case?
Use it to sanity-check your own figure, not to replace it. The published average comes from large communities in a survey several years old, and your cost depends on rent, turn scope, concession practice and vacancy days.