Tag: data analysis

Revenue per available space (RevPAS) is a facility’s parking revenue for a stated period divided by its available capacity over that same period. It is analogous to the hotel industry’s revenue per available room. RevPAS normalizes revenue for facility size and helps you investigate changes in pricing and use; it does not by itself prove why revenue changed. It needs revenue records; occupancy counts alone can’t produce it.

What RevPAS Measures

A parking space works like a hotel room or an airline seat: if it sits empty during a period, the revenue it could have earned in that period is gone. RevPAS measures how well each space earns, whatever the facility’s size.

Facility A earns $100,000 a month from 1,000 spaces: $100 per space. Facility B earns $50,000 from 250 spaces: $200 per space. A brings in twice the total revenue, but B earns twice as much from each space. If you’re deciding where to invest in better equipment or where to test a new rate, that difference matters more than the totals.

Two things must always accompany a RevPAS figure:

• The period. Per day and per month give very different numbers, and neither is comparable to the other.
• The capacity over that period. Use spaces that were actually available, not the painted total. If a level closes for half the month, count its spaces for the half when they were open. A single month-end count would misstate the denominator.

Without both, two RevPAS figures can’t be compared.

How to Calculate RevPAS

Before the step-by-step math, here is the whole idea in one picture. RevPAS breaks into two parts: how much of your capacity gets used, and how much each hour of use earns.

The basic calculation

Take a 300-space lot open 10 hours a day, every day (300 operating hours in a 30-day month). That schedule matters later: many municipal and commercial garages run 24 hours, so check your own facility’s hours before reusing these numbers. In the month, the lot takes in $45,000 from all sources: hourly parkers, monthly permits, and validations.

• Monthly RevPAS = $45,000 ÷ 300 = $150 per space per month
• Daily RevPAS = $150 ÷ 30 = $5.00 per space per day

That’s all a facility with unchanged capacity and operating hours needs for monthly reporting. When capacity changes, calculate available space-hours instead: sum the spaces open in each hour of the period.

Multiplying revenue per available space-hour by the scheduled open hours gives the same per-space figure as dividing by effective available spaces. For example, if 100 of 300 spaces close for half of a 300-hour month, available space-hours are (300 × 150) + (200 × 150) = 75,000, or 250 effective spaces. Revenue of $45,000 would be $0.60 per available space-hour, or $180 per effective space for the month. Report the closure and denominator alongside the figure.

Splitting RevPAS into revenue per occupied hour and utilization

When RevPAS changes, the next question is what moved: use of available capacity, revenue earned per occupied hour, or both. The trick is to measure both in the same unit: the space-hour, meaning one space for one hour. The second component reflects rates, permit mix, discounts, and payment capture together; it is not a pure price measure.

Return to the constant-capacity example: 300 spaces, open 10 hours a day throughout the month. In general form, the split is:

Plugging in the numbers:

1. Available space-hours = spaces × hours open. 300 × 10 hours × 30 days = 90,000.
2. Occupied space-hours = the total hours cars were parked, from counts or entry and exit records. Say 45,000.
3. Occupancy rate = occupied ÷ available space-hours = 45,000 ÷ 90,000 = 50%.
4. Revenue per occupied space-hour = revenue ÷ occupied space-hours = $45,000 ÷ 45,000 = $1.00. This is realized revenue per hour of use, after the effects of permits, discounts, validations, and the rate mix.

Multiply the two and you get revenue per available space-hour: 50% × $1.00 = $0.50. Multiply by the hours open to get back to the period figures:

• Per day: $0.50 × 10 hours = $5.00
• Per month: $0.50 × 300 hours = $150

These match the basic calculation, which is the check that the split is right. Now a change in RevPAS can be traced: if next month’s RevPAS rises, you can see whether occupancy rose, revenue per occupied hour rose, or both.

The mistake to avoid

A tempting shortcut is occupancy × average ticket. It does not account for how long each car stays or how many times each space turns over.

For a separate example, take a 300-space lot open 10 hours a day, with all revenue coming from parking tickets. It sells 600 tickets at an average of $2.50, earning $1,500 a day. If each car stays an average of 2.5 hours, occupied space-hours are 600 × 2.5 = 1,500 out of 3,000 available space-hours: 50% occupancy. Occupancy × average ticket gives 50% × $2.50 = $1.25, but actual daily RevPAS is $1,500 ÷ 300 = $5.00 per space. The shortcut mixes a proportion of available space-hours (occupancy) with a price per visit (ticket). It leaves out the two parking sessions per available space per day.

Hotels can use occupancy × average daily rate because a room is usually sold once per night. A parking space can be sold several times a day, which is why the calculation has to go through space-hours.

What Data You Need

• For RevPAS itself: revenue by period from every channel (hourly, monthly permits, validations, events), assigned to the right facility and period; plus available spaces and operating hours over that period, including partial closures.
• For the revenue-and-utilization split: occupied space-hours, from sufficiently frequent counts or from entry and exit records. Sparse snapshots cannot establish total occupied hours.
• A caveat for ticketless systems: with license plate recognition (LPR) or mobile pay-by-plate, occupancy and payment come from different records, and they can drift apart. If you measure occupied hours from paid sessions, vehicles that park without paying or registering are invisible, so occupancy looks lower and revenue per occupied hour looks higher than it is. If you measure from camera reads, those vehicles are counted but missed or misread plates add noise to dwell times. Either way, the gap between vehicles seen and vehicles paid is the payment-capture issue covered under What Moves RevPAS.

Occupancy data alone won’t get you there. In my analytics project on Istanbul’s municipal parking network, the city’s feed reports capacity and empty spaces for each facility, which supports occupancy. It has no revenue at all. So the project doesn’t calculate RevPAS or make any revenue claims, even though it would be easy to make up a plausible-looking number. The rule applies to any operator: don’t publish a metric your data can’t support.

Price vs. Occupancy: The Trade-Off

One reason to track RevPAS is to see the combined revenue effect when price and use both change. A higher rate can come with lower occupancy, but the response varies by facility and period. RevPAS shows the net revenue per available capacity; compare other conditions before attributing the change to price.

For this simplified example, assume each occupied space serves one paying car for the full operating day, with no turnover, permits, discounts, or unpaid parking. A 200-space lot charges $8 a day at 95% occupancy: 190 cars generate $1,520 a day, or $7.60 per available space. It raises the rate to $14. If occupancy falls to 70% under the same assumptions, 140 cars generate $1,960 a day, and daily RevPAS becomes $9.80 per space: 29% higher, even with about a quarter fewer cars. With shorter stays or turnover, use the space-hour calculation instead.

The 70% in that example is an assumption, and that’s the point to take from it. You won’t know how drivers respond until after the change. Some move to a nearby lot, some shift their arrival times, some pay. So:

• Compare RevPAS for several weeks before and after, against the facility’s own normal range and similar days or seasons. A change inside the range the facility usually varies over the past six to twelve months may be noise.
• Split the change into utilization and revenue per occupied hour, as above; then investigate what moved each component.
• If you run nearby facilities, watch whether demand simply moved next door.

What Moves RevPAS

• Pricing by time of day. Charging more during the hours a facility runs near capacity and less when it’s quiet. Set those hours from the facility’s own occupancy history rather than a fixed trigger.
• Event pricing. A flat rate during events, when demand is predictable and high.
• Reserved spaces. A named reserved space sits empty whenever its holder is away. Moving to unassigned permits lets others use it.
• Permit mix. How many spaces go to monthly permits and whether permit arrivals overlap with peak transient demand. Any decision to sell more permits than allocated spaces needs its own history and stress test.
• Payment capture. If 80% of parked vehicles pay and that rises to 95%, revenue from those vehicles rises 18.75% (95 ÷ 80 = 1.1875), assuming the number of parkers and the average payment per paying vehicle stay unchanged. That average depends on rates, length of stay, and discounts; the percentage is not guaranteed if any of those change or drivers leave because enforcement tightened. Measuring the change needs vehicle counts matched to payments.

A Property-Value Sensitivity Illustration

For owners, sustained revenue changes may affect net operating income (NOI), which is one input to property valuation. The following calculation shows sensitivity to assumptions; it does not estimate a realizable sale price.

A 400-space garage raises monthly RevPAS by $25, from $125 to $150. That’s $10,000 a month, or $120,000 a year. Assume 85% of the added revenue reaches NOI after added costs: $102,000. At an assumed 7% capitalization rate, $102,000 ÷ 0.07 is about $1,457,000 of added value.

That’s arithmetic, not an appraisal or evidence that a buyer would pay $1.457 million more. Actual value depends on the lease or management structure, costs, market, appraisal method, and whether the higher RevPAS is expected to last.

A Monthly RevPAS Check

1. Calculate RevPAS for each facility for the month. State the period and the space count.
2. Compare it with that facility’s last six to twelve months. Judge the change against how much it normally varies.
3. If it’s off track, split it. Did occupancy move, or revenue per occupied hour?
4. Assign one action, with an owner and a date.

How to build a monthly KPI scorecard shows a spreadsheet layout for tracking RevPAS alongside targets and variance.

Frequently Asked Questions

What is a good RevPAS?

There’s no general figure. RevPAS depends on location, facility type, pricing, and demand. Compare each facility with its own history and with your other facilities.

What’s the difference between RevPAS and average ticket?

Average ticket is revenue per parking session. RevPAS is revenue per available space. A higher average ticket can come with lower RevPAS if fewer cars park.

Should RevPAS include monthly permit revenue?

Yes. Include all parking revenue for the facility. Tracking the hourly and permit shares separately helps explain changes.

Can I calculate RevPAS from occupancy data?

No. RevPAS needs revenue. Occupancy data tells you how much of the capacity was used, not what it earned.

Parking occupancy is the number of occupied spaces divided by the number of usable spaces at a specific moment, times 100. The formula is the easy part. Before you trust the result, confirm what your capacity and available-space numbers mean, record when each reading was taken, reject invalid records, and collect readings over time. One reading describes one moment; it cannot establish the busiest time of a day or week. Manual counts, gate and payment records, and sensor or feed data can all help if you check their scope and limitations.

Three Different Numbers People Call “Occupancy”

When someone says a lot is “at 60%,” they could mean one of three things:

• Occupancy at a moment: occupied spaces ÷ usable spaces at one time. “At 10:15 Tuesday, 180 of 200 spaces were taken: 90%.”
• Observed peak occupancy: the highest reading in your sample across a defined period, such as a day or a week. Report the count times and number of readings; sparse sampling can miss the true peak.
• Average utilization: the share of available space-hours that were used over a period. A 100-space lot open 10 hours has 1,000 space-hours available. If cars occupied 600 of them, utilization is 60%.

These answer different questions. Utilization tells you how much of your capacity earns its keep over the day. Peak tells you whether you ran out of room.

Mixing them up is the most common occupancy mistake. A garage reports 45% occupancy to its owner, a 24-hour average. It was effectively full from 11:30 to 1:30 every weekday, turning drivers away. Both statements are true. Only one of them helps anyone decide whether to add permits, change prices, or send overflow elsewhere.

The Formulas

Occupancy % = occupied spaces ÷ usable spaces × 100, measured at a stated time.

Three details decide whether that number is right.

Use a capacity denominator that matches the spaces being counted. Design capacity is every striped space. Usable capacity excludes spaces closed for repairs or blocked by construction or snow. If you report general-access parking separately, also exclude reserved, accessible, or loading spaces from both the numerator and denominator for that specific pool; do not drop them from a total-facility occupancy figure. A 200-space garage with a closed level of 40 spaces that fills its remaining 160 is 100% full, not 80%. State which capacity you used.

Deriving “occupied” from “available” works only if both describe the same spaces at the same time. Many systems report empty spaces rather than occupied ones. Occupied = capacity − available is correct only when the capacity and the available count cover the same set of spaces and were recorded at the same moment.

For gate or entry/exit systems, count the accumulation. Occupancy at a given time = the starting count + entries − exits. If 40 cars are inside at 6:00 AM, 180 enter and 95 leave by 10:00 AM, accumulation is 125 cars inside at 10:00. Divide that by usable capacity to get the occupancy rate. Accumulation drifts over time (more on that below), so it needs a regular reset against a physical count.

Before You Calculate: Check What Your Data Means

I’m building an analytics project on Istanbul’s municipal parking network, which publishes a live feed of its facilities with capacity and empty-space counts. Before calculating a single occupancy figure, I audited what the feed actually returns. The arithmetic turned out to be the least of it.

Here’s a real reading from one snapshot on 2026-08-22. Facility 3068, an enclosed garage listed as open 24 hours, reported a capacity of 1,029 and 589 empty spaces at 02:31 Istanbul time. That’s 1,029 − 589 = 440 occupied, or 42.8%.

The calculation took one line. The audit found five things that would have made numbers like that wrong without anyone noticing:

1. Invalid requests return a normal-looking record. Asking for a facility ID that doesn’t exist returns a record with a capacity of 1 and 1 empty space instead of an error. Included in a calculation, it looks like a tiny, empty lot. Lesson: define what a valid record looks like and reject the rest before you calculate.
2. The facility list has no timestamp. The list doesn’t say when each reading was taken. Lesson: if the source doesn’t timestamp a reading, record the time you retrieved it. A reading you can’t place in time can’t be part of a peak.
3. A status field has no documentation. The isOpen field is 0 for most facilities, including facility 3068, which lists 24-hour operation. Lesson: don’t guess what an undocumented field means. Leave it out until you can confirm it.
4. A time field has no timezone. The detail record’s update time doesn’t say which timezone it’s in. Lesson: confirm timezones before comparing readings from different sources.
5. There’s no history. The feed shows only the current state. Lesson: covered in its own section below.

Your data will have different quirks, but the questions carry over to any lot, garage, or system:

• What exactly does “capacity” include? Does it change when spaces close?
• What does “available” or “empty” mean, and does it cover the same spaces?
• Where does each reading’s timestamp come from, and in what timezone?
• What does an invalid, closed, or offline record look like?
• Does the source keep history, or only the current state?

The project’s code and evidence are public at github.com/johnserra/istanbul-parking-analytics. These figures come from a single audit snapshot. They aren’t a finding about how full Istanbul’s garages are.

Istanbul parking data source: Istanbul Metropolitan Municipality (IBB) Open Data Portal, 2026-08-22 audit snapshot. The IBB Open Data License v1.0 requires this attribution: “Contains public sector information licensed under the Attribution 4.0 International (CC BY 4.0).”

Three Ways to Collect Occupancy Data

Manual counts

Someone walks the facility on a schedule and counts occupied spaces. It costs staff time and nothing else, and it’s the most direct measurement there is.

Choose count times from how the facility is used. An office garage, a retail lot, a hospital, and an event venue peak at different times, and a single universal schedule will miss some of them. Count at the times you expect to be busiest, plus a quiet period for comparison, on the days that matter (weekdays, weekends, or both). Record each count with its date and time on a simple sheet.

Gate, ticket, and payment records

If your facility has gates, ticketing, or pay-by-plate, you can reconstruct occupancy from entries and exits, usually in 15- or 30-minute intervals. The data already exists, so this is often the best place to start before buying anything new.

The catch is drift. Tailgating, unreadable tickets, gate arms left up, and cars that exit without being recorded all push the running count off over time. Reset it against a physical count at a known quiet point on a regular schedule, and compare the two to see how far it drifted.

Sensors, cameras, and live feeds

Per-space sensors, entry counters, camera counts, and live data feeds give you frequent readings without anyone walking the lot. They’re worth it when a decision needs real-time information, such as guidance signs showing available spaces or prices that change by time of day.

Automation doesn’t skip the checks above. A sensor feed has its own capacity definition, its own offline states, and its own timestamps, and all of them need confirming. For planning questions that do not need a live feed, start by testing whether counts and existing records answer the question before paying for hardware.

One Reading Isn’t a Peak: Build Your Own History

The Istanbul feed only reports the current state, so my project stores a snapshot each time it reads the feed. Without that, there’s no history. The same is true of most live feeds and many gate-system dashboards: they show you now, and “now” disappears.

Two rules from the project carry over:

• Keep the raw readings, not only the calculated percentage. If you later find that a capacity figure was wrong or an invalid record slipped through, you can recalculate.
• Match the claim to the history you have. My project won’t forecast occupancy or trigger capacity alerts for a facility until it has 26 weeks of readings with at least 90% coverage. That’s a high bar for forecasting. For simple peak reporting, the practical version is to say what the peak is based on: “Peak of 96% at 10:30 on Tuesdays, from 18 weekday readings over three weeks” is an honest statement. “Peaks at 96%” from one busy morning isn’t.

What Counts as “Full”?

You may see 85% cited as a parking occupancy target. That number comes from a point-in-time curbside-parking context, not a universal goal for lots or garages. For example, SFMTA’s SFpark policy describes a commonly cited 85% curbside threshold at a single moment while using a 60–80% average occupancy target across a longer period to keep spaces available on each block. Those are different measurements and settings; neither gives your facility its own “full” threshold. Practitioners have also questioned relying on any occupancy target alone: in a May 2026 Parking Today piece, Cole Jaillet argues that occupancy is a snapshot rather than a behavior, and that two blocks at the same percentage can serve very different numbers of vehicles depending on how long each stays.

For a specific facility, set two thresholds from its own history: a “busy” level where drivers start having trouble finding a space, and a “full” level where the facility is effectively out of room. Look at the readings from times you know were difficult (complaints, turned-away drivers, staff reports) and see where occupancy stood. Then report how long the facility stayed above each threshold, not only whether it crossed.

A Two-Week Occupancy Audit

If you’re starting from nothing, this gives you a defensible baseline:

1. Set usable capacity and write down exactly what you counted and excluded.
2. Choose count times from the facility’s use pattern. Cover weekdays and weekends if both matter.
3. Record every count or snapshot with its timestamp in one sheet. Keep the raw numbers.
4. Reconcile counts with transaction and permit records for the same times. Account for vehicles already present, exits, permits, validations, and unpaid sessions before treating a gap as a counting or payment problem.
5. Report peak, time near capacity, and how many readings each is based on. End with one follow-up action, an owner, and a date.

Once occupancy is reliable, it can support revenue analysis when you also have revenue records for the same facility and period. Occupancy data alone cannot show what a space earned. A full lot of monthly permit holders and a lot turning over transient drivers can look identical on an occupancy chart while producing very different revenue, which is why revenue per space needs its own treatment.

Frequently Asked Questions

What is the formula for parking occupancy rate?

Occupied spaces ÷ usable spaces × 100, measured at a stated time. Use the spaces actually available at that moment, not the design total.

How often should parking occupancy be measured?

Often enough to sample the periods when your facility is likely busiest. For manual counts, that may mean several readings across the expected peak. Automated intervals should match the decision you need to make and the feed’s actual update frequency. State how many days and readings your observed peak covers.

Is 100% occupancy good?

Not usually. A lot at 100% is turning drivers away and has no room for permit holders who arrive late. How far below full a facility should run depends on its layout and its customers.

Can I measure occupancy without sensors?

Yes. Manual counts and gate or payment records answer most planning questions. Sensors earn their cost when a decision needs real-time data.

Track six metrics: peak occupancy, hours near capacity, revenue per available space (RevPAS), the split between hourly and monthly revenue, operating cost per space, and payment capture rate. Most KPI lists skip a practical point: each metric needs specific data. Occupancy counts support only the first two. RevPAS and the revenue split need revenue records, cost per space needs expenses by facility, and payment capture needs vehicle counts matched to payments. Start with the metrics your data supports.

This guide covers what each metric tells you, what data it takes, and how to review them without adding reports nobody reads.

Why Total Deposits Hide What’s Happening

Many lots are run from one number: how much was deposited this month. It’s the number that matters most in the end, and on its own it explains almost nothing.

A deposit total can’t tell you whether revenue rose because rates went up or because more cars parked. It can’t separate hourly parkers from monthly permit holders. And it can’t show the Tuesday mornings when the lot was full, and drivers went somewhere else. An empty space during a busy hour is revenue that can’t be recovered later, and a full lot turning drivers away is revenue you never see at all.

The fix isn’t a long list of metrics. Keep the scorecard to four to seven numbers, depending on the business, and cut any metric that doesn’t move the needle. For parking, each metric should lead to a decision about staffing, pricing, permit allocation, or enforcement. If nobody would change anything when a number moves, it doesn’t belong. How many KPIs a small business should track makes the general case.

Start With the Data You Have

Before choosing metrics, check what your data can support. I learned this building an analytics project on Istanbul’s municipal parking data. The city’s parking API reports capacity and empty spaces for about 250 facilities, which is enough to calculate occupancy. It has no revenue, no payments, and no record of when individual cars arrive or leave. So the project leaves out revenue per space, payment compliance, and how long cars stay. Those metrics could be calculated from other data, but not from that feed, and a number the data can’t support is worse than no number: it looks precise and is wrong.

The same test applies to any lot. Here’s the scorecard with the data each metric needs:

Go down the “data it needs” column and mark each row: have it, could get it, or don’t have it. That gives you your starting scorecard, and a short list of what to start collecting.

Istanbul parking data source: Istanbul Metropolitan Municipality (IBB) Open Data Portal, 2026-08-22 audit snapshot. The IBB Open Data License v1.0 requires this attribution: “Contains public sector information licensed under the Attribution 4.0 International (CC BY 4.0).”

Metrics 1 and 2: Peak Occupancy and Hours Near Capacity

Occupancy is the share of usable spaces that are occupied at a given moment. Averaged over a day, it hides the pressure that matters. A lot averaging 55% can be completely full from 9 to 11 every weekday morning, and those two hours are when drivers get turned away and permit holders complain.

So report two numbers instead of an average:

• Peak occupancy: the highest reading in each day or week.
• Hours near capacity: how many hours the lot stayed at or above your “effectively full” level.

Peak tells you whether you ran out of room. Hours near capacity tells you how long, which is what pricing and permit decisions depend on. One tight hour on Fridays is a different problem from four tight hours every weekday.

Getting these right depends on counting properly: using usable capacity rather than the number on the sign, timestamping every reading, and building enough history to describe an observed peak and its sampling limits.

Metric 3: Revenue per Available Space (RevPAS)

RevPAS is revenue divided by available spaces for a stated period, such as a day or a month. It does two things occupancy and revenue can’t do alone.

First, it makes facilities of different sizes comparable. A 40-space lot and a 400-space garage can be judged on the same scale.

Second, it normalizes revenue across available capacity. A lot raises its hourly rate from $5 to $10, and occupancy during those hours falls from 80% to 30%. Measured per space-hour, which is RevPASH (revenue per available space-hour, or occupancy × rate), revenue falls from $4.00 (0.80 × $5) to $3.00 (0.30 × $10). Total revenue over equal hours and capacity would show the decline too; the normalized measure helps compare different facilities or periods. RevPAS is the same idea over a longer period, such as a day or a month. The before-and-after comparison alone does not prove the rate change caused the decline. Before blaming the price, check for events, weather, and day-of-week patterns in the same hours, and for nearby lots’ occupancy if you can get it.

RevPAS needs revenue by period and capacity over that same period. If spaces open or close during the period, use available space-hours before converting the result into a per-space figure; a single end-of-month space count would distort the comparison.

Metric 4: The Mix Between Hourly and Monthly Parkers

Most lots serve two kinds of customers, and they pay differently.

• Monthly permit holders bring predictable revenue that arrives on schedule. The trade-off is a lower rate per hour of use, and they hold spaces during the busiest hours.
• Hourly (transient) parkers usually pay more per hour, but demand swings with weather, events, and the day of the week.

The mix is a decision, not an accident. Track the share of revenue from each, and, if you can, the share of peak-hour spaces each uses.

Overselling permits

Permit holders may not all park at the same time. Some work from home, travel, or leave early. An operator may consider selling more permits than the spaces set aside for them, but that decision needs counts and a plan for days when more holders arrive than expected.

Set the oversell level from your own counts, not from a ratio you read somewhere. Count how many permit holders are present at the busiest time, over several weeks, and use the highest share you see.

Here is how it works for a 400-space garage that sets aside 250 spaces for monthly permits:

• Observed peak: over six weeks of counts, the highest simultaneous presence was 75% of permit holders.
• Buffer: the operator chooses a 10% space buffer, so modeled peak permit parking stays at or below 225 spaces.
• Permit cap: 225 ÷ 0.75 = 300 permits, or 120% of the 250 spaces.
• Left for hourly parkers: the other 150 spaces.

A future day can exceed the observed rate, so this calculation is a scenario to stress-test against unusual days and contractual obligations, not a guarantee of space.

Recheck the counts regularly. If more permit holders start coming in every day, the safe number of permits drops.

Metric 5: Operating Cost per Space

Add up what it costs to run each facility: labor, payment processing, equipment and software, utilities, insurance, sweeping, snow removal, and repairs. Divide by the number of spaces for the period.

Cost per space lets you compare facilities and spot drift. If one garage costs noticeably more per space than your others, or more than it did last year, find out why before the next contract renewal. The comparison that matters is against your own history and your own other facilities, since a generic industry ratio won’t reflect your labor market, climate, or contract terms.

This needs expenses recorded by facility, not lumped together. If your books combine several lots, splitting them is the first job.

Metric 6: Payment Capture Rate

Payment capture rate is the share of parked vehicles that have paid or been validated. Unpaid parking is lost revenue, and it hides inside an occupancy count, because an unpaid car and a paid car take up the same space.

Measure it with spot checks: count the vehicles in the lot, then match them against active payments, permits, and validations at the same moment. Paid ÷ parked is the rate. Repeat at different times and days, since compliance often varies.

This metric needs matched payment data. Occupancy counts alone can’t show it, which is exactly the kind of metric to leave off until you can support it. There’s no universal target. Set one from your own baseline, and when it drops, check signs, payment machines, and apps before assuming drivers are avoiding payment.

The spot-check rate measures compliance at that moment. If you also issue notices or invoices to non-payers, track separately what share of them is eventually paid. A lot with a low capture rate and a high recovery rate has a different problem from one that is low on both. Whether and how you can pursue unpaid sessions depends on local law and regulation, so treat recovery as its own measure and don’t fold it into the capture rate. Lots that do not use a modern tech stack have no session records to match, so for them the spot check is the only measure available.

A Weekly Review in Four Steps

Once the scorecard is set up, a short weekly review keeps it useful:

1. Occupancy: peak and hours near capacity by facility. Any lot running full for longer than usual?
2. RevPAS: compared against each facility’s own trailing baseline. Set how much movement counts as normal based on how much the number has actually varied over the past several months.
3. Mix and capture: permit share, any oversell pressure, and the latest spot check.
4. Actions: for each facility that’s off track, one action, one owner, and a date.

The last step is what turns the review into management. How to build a monthly KPI scorecard shows a spreadsheet layout with targets, variances, and status colors that works for a set of lots as well as for a whole business.

Parking KPI Checklist

• I know which of the six metrics my current data supports.
• I report peak occupancy and hours near capacity, not just averages.
• I calculate RevPAS using spaces that were actually available.
• I track revenue from hourly and monthly parkers separately.
• My permit oversell level comes from my own counts.
• Expenses are recorded by facility.
• I run payment spot checks at different times.
• Each weekly review ends with named actions and dates.

Frequently Asked Questions

What is the most important KPI for a parking lot?

It depends on what you’re deciding. For pricing and comparing facilities, RevPAS, because it combines price and use. For capacity decisions, peak occupancy and hours near capacity.

What is a good occupancy rate for a parking lot?

It depends on the facility and on what “full” means for it. Set the threshold from its own layout, operating history, and instances when drivers struggled to find a space.

How often should parking KPIs be reviewed?

Weekly for occupancy, mix, and capture, since they change quickly and the fixes are operational. Monthly for RevPAS and cost per space, once revenue and expenses are closed for the month.

What about turnover?

If you have entry and exit data, turnover (parking sessions per space per day) shows how intensively spaces are used, and it’s useful for lots that serve short visits. It needs session data that many lots don’t collect, so it’s an optional seventh metric.

Data analytics for a property management company means tracking a handful of operating numbers that can affect net operating income (NOI), rather than reading the monthly accounting package after the fact. Five starting measures are physical occupancy alongside rent collected against potential rent, work-order resolution, days vacant between tenants, rent more than 30 days late, and controllable operating expenses per square foot. Choose the ones that fit your portfolio and data; some inputs may already be in your property management software.

Accounting Reports Tell You What Happened. Operating Metrics Tell You What's Coming.

Most property management systems produce a thorough monthly package: rent roll, income statement, general ledger, aged receivables. It's accurate, and it's backward-looking. By the time a slow repair or a rising utility bill shows up on the income statement, the tenant is already frustrated or the money is already spent.

Operating metrics are the leading indicators behind those accounting results. Repair times, turn times, and late balances move weeks or months before NOI does, which gives you time to act.

The test for including any metric is simple: does it help the property earn more revenue, run more efficiently, or cost less to operate? If it doesn't, leave it off. Keep the scorecard to four to seven numbers so someone actually acts on each one; how many KPIs a small business should track explains why that range holds.

1. Physical Occupancy and Rent Collected Against Potential Rent

Physical occupancy is the share of leasable space that has a tenant in it. For this scorecard, use a cash-collection ratio to show the share of potential rent actually collected. Some property reports call a related measure economic occupancy, but definitions vary, so keep the numerator and denominator visible:

Cash-collection ratio = rent collected for the period ÷ gross potential rent for the same period × 100

Here, gross potential rent is what the property would bring in if every leasable space were occupied at the documented market-rent assumption for that space and every tenant paid in full. Document that rent basis and use it consistently; scheduled rent in existing leases, billed rent, concessions, and cash collected are different amounts.

The two numbers can diverge because of vacancy, free-rent concessions, discounts, and unpaid balances. The gap is not all collectible debt: some of it reflects deliberate lease terms or empty space.

A building is 95% physically occupied. After two months of free rent on a new lease, one tenant well behind on payments, and a discounted renewal, it collects 84% of gross potential rent. The 11-point gap calls for a breakdown of vacancy, concessions, discounts, and overdue rent. The late balance is the part collections can pursue without changing lease terms.

2. Work Order Resolution Time

How quickly a routine repair gets fixed is something every tenant experiences, every month, long before the renewal conversation. Track three numbers:

• Time to first response: hours from the request to someone acknowledging it or being assigned.
• Time to complete: days from request to closed work order, for routine requests.
• Open requests past the property's service target: the backlog that needs someone's attention. Set that target by request type and urgency rather than using one cutoff for every repair.

Watch the trend rather than the average alone. A property whose completion time creeps up quarter after quarter has a staffing, vendor, or parts problem building, even if no single request looks bad. Slow repairs can also weigh on whether a tenant renews, so watch the renewal rate next to these numbers; it connects service back to NOI.

3. Days Vacant Between Tenants

Rent lost while a space sits empty doesn't come back. Measure the total days from move-out to the new tenant's lease start, and split it into three stages:

1. Move-out inspection and scoping: from the keys being returned to a defined list of work.
2. Make-ready: repairs, paint, flooring, cleaning.
3. Leasing: from rent-ready to a signed lease and move-in date.

The split tells you where time goes. If make-ready is the long stage, the question is vendor scheduling. If spaces sit rent-ready for weeks, it's a leasing or pricing question. A shared board listing every vacant space with its current stage and days in that stage is often enough to show the bottleneck.

4. Rent More Than 30 Days Late

A late balance is much easier to collect while it's young. The metric is the over-30-day late rate:

Over-30-day late rate = rent balances more than 30 days past due ÷ rent billed for the month × 100

Don't wait for the month-end aging report to find it. Send reminders and talk to tenants before balances become overdue. Use the over-30-day bucket as a review trigger; a balance approaching 60 days calls for a more direct follow-up under the lease and collection process. What counts as a normal level depends on the property type and tenant mix, so compare each property against its own history rather than a generic benchmark.

5. Controllable Operating Expenses per Square Foot

Some operating costs, like property taxes and insurance, are set outside day-to-day management. Others, like janitorial, landscaping, repairs, and utilities, are controllable. Track those separately and divide by rentable square feet so you can compare properties of different sizes and each property against its own past months.

Unexplained changes are the trigger. Water cost per square foot at one building runs 40% above the same month last year with no change in occupancy. That's a reason to walk the building before the next bill: a running toilet, an irrigation zone stuck on, or a meter reading that needs checking. The metric doesn't diagnose the cause. It tells you where to look.

How to Start Without Buying New Software

You probably have all five metrics' raw data already.

1. Export what you have. Rent roll, aged receivables, work order history, and the general ledger by property are standard exports from most property management systems. Put them into one spreadsheet, one tab per export.
2. Build one screen. Five metrics, one row per property, each with a target, the current value, and a green, yellow, or red status. How to build a monthly KPI scorecard shows the layout.
3. Review it weekly or monthly with the people who can act. The review is a short conversation about what the numbers mean, who will act on them, and by when. How to get your team to actually use your reports covers running that meeting.

Stay in the spreadsheet while it gives the team a reliable, timely view. If recurring exports and reconciliation consume more staff time than a reporting tool would save, or multiple systems make errors hard to catch, evaluate an upgrade against those actual costs.

If your portfolio includes parking, its operating metrics need their own scorecard. The KPIs every parking lot operator should track covers capacity, utilization by hour, revenue per space, and payment capture, which may matter there alongside property-level measures.