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:
| Metric | How to calculate it | Data it needs | Decision it drives |
|---|---|---|---|
| Peak occupancy | Highest occupied spaces ÷ usable spaces in a period | Timestamped counts or sensor readings | Permit allocation, whether to add or share space |
| Hours near capacity | Hours at or above your “full” threshold | The same counts, taken at regular intervals | Pricing by time of day, overflow plans |
| RevPAS | Revenue ÷ available spaces, for a stated period | Revenue by period; count of spaces actually usable | Pricing, allocation, comparing facilities |
| Hourly vs. monthly mix | Share of revenue (and of peak spaces) from each type of parker | Revenue by channel; permit list; ideally counts by parker type | Permit limits, pricing |
| Operating cost per space | Operating expenses ÷ spaces, per period | Expenses recorded by facility | Staffing, automation, contract terms |
| Payment capture rate | Paid vehicles ÷ parked vehicles in spot checks | Vehicle counts matched to active payments at the same moment | Enforcement, signs, fixing payment problems |
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:
- Occupancy: peak and hours near capacity by facility. Any lot running full for longer than usual?
- 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.
- Mix and capture: permit share, any oversell pressure, and the latest spot check.
- 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.