Public Sector Performance Dashboard
A nursing graduate has a hospital job lined up for July. She sent her license application in May. It's late June, the online status still says "in review," and she's on hold with the licensing agency for the third time this week. Inside the agency, nobody thinks there's a problem. Last month's report showed the backlog up a little. So did the one before.
Licensing reaches a lot of working people. In 2025, 35.3 million employed Americans held a license, 21.6% of everyone in work, according to the Bureau of Labor Statistics. BLS defines a license as one awarded by a government agency, so each slow month in a licensing office is a delayed start date somewhere else.
The staffing behind that work is tight. In NASBO's Spring 2026 Fiscal Survey of States, governors' recommended budgets for fiscal 2027 had median general fund spending growth of 0.6%, and 14 states proposed eliminating vacant positions. Hiring hasn't stopped: state and local governments outside education still had 390,000 job openings in August 2026, a 3.8% openings rate in the preliminary BLS estimate. But a seat left open in a licensing unit doesn't hurt on day one. It hurts slowly.
That's the story this board tells. The agency is invented, a state licensing department, and its monthly figures are modeled, not taken from a real agency. City services like 311 and potholes live on the local government example; this one is about applications and caseloads. None of the ten targets has a published benchmark for state licensing work, so all of them are starting points. Where a definition borrows from federal practice, the page says which.
What is a Public Sector KPI Dashboard?
Overview
For a state agency that processes applications, a public sector KPI dashboard shows how long people wait for a decision, how much of the queue has waited too long, and whether the staff and money behind the service are keeping up.
It starts at the front door. Applications received sets the workload. Online application share shows how much arrives through the portal, where the agency can check documents before anything is submitted. First-time-right rate shows how many applications go through without a deficiency letter, the biggest single source of rework in licensing.
Then the queue. Median processing time is the number the public and the legislature see. The backlog, applications open more than 30 days, is the aged end of the queue, and it moves before the median does.
Then the people waiting. Contact center wait time rises with the backlog, because most calls are some version of "where's my license?" Customer satisfaction, asked right after the transaction, catches up a month or two later.
Underneath sit three inputs: staff vacancy rate, cost per application processed and the share of improvement projects on schedule. Between them they explain most of the movement above.
Who benefits from this dashboard?
- Agency Director: Median processing time and satisfaction are what the governor's office and legislators ask about. The vacancy rate on the same screen is usually the answer to their next question.
- Licensing Operations Manager: Backlog and first-time-right rate show where the queue is stuck, whether that's missing documents, a board that meets monthly, or too few reviewers.
- Contact Center Supervisor: Wait time read against the backlog tells them if a bad week is a gap on the phones or a queue problem upstream that no amount of phone cover will fix. The customer service example goes deeper on the phone side.
- Budget Officer: Cost per application against the same month last year, beside the vacancy rate, shows whether a saving is real or just deferred.
- Project Management Office: Projects on schedule is where a backlog push shows up, because the first people borrowed for the queue are usually the analysts.
Start the processing-time clock on the day an application arrives, not the day it's complete. An agency that counts from "complete" can make its median fall by sending more deficiency letters, which adds weeks for the applicant and nothing to the board.
Dashboard Example Overview
The agency licenses about 40 occupations, from registered nurses and electricians to cosmetologists and real estate brokers, and handles around 124,000 new and renewal applications a year. Its licensing division has 124 authorized positions, a contact center and an online portal. It reports monthly against ten KPIs:
- License Applications Received: New and renewal applications in the month, every channel. It's the one volume figure on the board, so it has no target. It peaks each summer with new graduates.
- Online License Application Share: Applications received through the portal divided by all applications received. A portal can refuse to submit without the required documents. A paper form can't.
- License Application First-Time-Right Rate: Applications decided without a deficiency notice or a return to the applicant, divided by applications decided.
- Median License Processing Time: Median calendar days from receipt to decision for applications decided in the month. It's the convention USCIS uses for its historical processing times: the time it took to complete half the cases.
- License Application Backlog: Applications still open more than 30 days after receipt, at month end. The 30 days is this agency's own standard. The split between backlog and inventory follows the Department of Veterans Affairs, which counts as backlog only the rating claims pending more than 125 days.
- Licensing Contact Center Wait Time: Average speed of answer: total minutes callers spent in queue divided by calls answered.
- Licensing Customer Satisfaction: Post-transaction survey respondents answering 4 or 5 out of 5 to "I am satisfied with the service I received," the standard statement in OMB Circular A-11 Section 280 (2022 edition), divided by all respondents. The federal guidance doesn't bind a state, but the same wording keeps the score comparable.
- Cost per License Application Processed: Total division cost for the month (staff, overtime, systems and overhead) divided by applications decided. It swings with the seasons, so it's read against the same month last year.
- Licensing Staff Vacancy Rate: Vacant authorized positions divided by authorized positions, at month end. Authorized, not funded, or a frozen post disappears from the count.
- Agency Projects On Schedule: Active improvement and IT projects with schedule variance under 10%, divided by active projects. The 10% line is the green threshold on OMB's federal IT Dashboard, as GAO described in 2016.
The two years are about open seats, and what they cost before anyone put a price on them.
October 2024 looks healthy. The backlog is 1,240 applications, the median decision takes 14 days and 75.4% of surveyed applicants are satisfied. Nine of the 124 positions are empty, a 7.3% vacancy rate.
Then the backlog rises in January 2025, and again in each of the next ten months. Never by much: 52 applications in January, 62 in February, 54 in March. The biggest jump is 114, in June, and the monthly review puts it down to the graduation rush. The median creeps up a day at a time, from 14 in January to 20 in July. Each month's pack compares the month with the one before, and each time the change looks like noise.
The vacancies drift the same way. Retirements and resignations aren't backfilled straight away, because posting, interviews and background checks take months, and nobody pushes, because open seats show up as savings. By October 2025, 18 of the 124 positions are empty, 14.5%. Meanwhile cost per application looks better than ever. July 2025 comes in at $101.99, the lowest on the board at the time, because summer volume was high and the payroll was short.
In October 2025 the budget analyst preparing the fiscal 2027 request needs a chart for the legislative hearing. She charts twelve months of backlog with the vacancy rate beneath it, and the two have the same shape. The backlog is up 62%, from 1,240 to 2,004. The median is 22 days. The average caller now waits 7.8 minutes, up from 5.4, and satisfaction has slid to 68.1%. Nothing broke. It just never stopped drifting.
The fix is unglamorous. HR moves licensing specialist roles to continuous posting with a standing list of qualified candidates, and new reviewers start in pairs from December. Reviewers work limited overtime from November to March. In January 2026 the portal stops accepting applications with documents missing, and supervisors work the over-30-day list oldest first every Monday. Two business analysts from the portal project move onto the queue for the winter.
The recovery is as slow as the drift. The backlog peaks at 2,060 in November 2025, falls by 12 in December, then by 66, 101, 109 and 117 over the next four months. First-time-right climbs from 70.9% in November 2025 to 81.3% in September 2026, and online share reaches 78.4%. By September 2026 the backlog is 1,294, the median is back to 14 days, callers wait 5.5 minutes, and the vacancy rate is down to 8.9%, still 11 open seats.
Some numbers went the wrong way, and the fix caused most of it. Cost per application hit $140.82 in December 2025, against $135.80 a year earlier, with overtime on top and new hires paid while they trained. September 2026 is $113.35, still above the $108.92 of September 2025. That's the price of the people the 2025 figure said weren't needed. Projects on schedule fell from 83.3% in October 2025 to 58.3%, 7 of 12, from January to March 2026 while the analysts were on loan, before recovering to 90.0%. And satisfaction, at 74.5%, still hasn't reached the 75.4% where it started. Applicants who waited in 2025 remember.
Creating a Public Sector Dashboard in SimpleKPI
-
Step 1 Draw the backlog line at your published service standard
Write down the four dates every application carries: received, complete, decided and issued. Then define the backlog by age, with a threshold that matches the service standard you publish, and keep applications waiting on the applicant in the count, flagged, or the backlog can be shrunk by mailing letters. Two units counting the same caseload different ways can report medians a week apart.
For testing a measure before you commit to it, the KPI selection guide goes further.
-
Step 2 Put licensing, the phones, HR and finance on one source list
The licensing system owns applications, backlog, median and first-time-right. The phone platform owns wait time, the survey tool owns satisfaction, HR's position control owns vacancies, finance owns cost and the project office owns schedule status. One source per number, written down, so the vacancy rate on the board is the same one HR gives the budget office.
-
Step 3 Put twelve months on every chart the director sees
Month-on-month comparisons are where drift hides. Supervisors need a weekly look at the backlog by age band. The director's monthly review needs a 12-month trend on every chart, with the target line drawn in. The budget office and the legislative hearing need the same month last year. This guide to writing a KPI report covers how to lay that out for readers who skim.
-
Step 4 Price every vacancy in backlog days
Before an open position is booked as a saving, show what it's doing to the backlog and the median. Pair cost per application with processing time, so a cheap month that built a queue is visible as one. Cost per application and first-time-right are both ratios of counts the agency already keeps; this guide to building KPIs with formulas shows how to calculate them from those counts.
List of Public Sector KPIs
None of these ten has a published benchmark for state licensing work, so every target below is this example's starting point. Replace each one with a figure from a year of your own history.
| KPI Name | Typical Target | Description / Calculation |
|---|---|---|
| License Applications Received | No target (volume) | New and renewal applications received in the month, all channels. |
| Online License Application Share | ≥ 75% (starting point) | Applications received through the portal divided by all applications received. |
| License Application First-Time-Right Rate | ≥ 80% (starting point) | Applications decided with no deficiency notice or return to the applicant divided by applications decided. |
| Median License Processing Time | ≤ 15 days (starting point) | Median calendar days from receipt to decision for applications decided in the month. Lower is better. |
| License Application Backlog | ≤ 1,200 (starting point) | Applications open more than 30 days after receipt at month end. Lower is better. |
| Licensing Contact Center Wait Time | ≤ 5 minutes (starting point) | Total minutes callers waited in queue divided by calls answered. Lower is better. |
| Licensing Customer Satisfaction | ≥ 75% (starting point) | Respondents answering 4 or 5 on the A-11 satisfaction statement divided by all respondents. |
| Cost per License Application Processed | ≤ $115 (starting point) | Total division operating cost divided by applications decided. Compare with the same month last year. |
| Licensing Staff Vacancy Rate | ≤ 8% (starting point) | Vacant authorized positions divided by authorized positions at month end. Lower is better. |
| Agency Projects On Schedule | ≥ 80% (starting point) | Active projects with schedule variance under 10% divided by active projects. |
Why Public Sector Dashboards Matter
In a state agency, the budget and the service are reviewed by different people, at different times, from different reports. Finance sees a payroll line coming in under budget. Operations sees a queue getting longer. Neither report says they're both describing the same open seats.
Several of these ten pull against each other. A rising vacancy rate lowers cost per application for months before it raises processing time, which is exactly how the drift in this example stayed hidden. First-time-right can be pushed up by refusing incomplete applications at the portal, which helps the queue but can send more people to the phones, so wait time needs watching while it climbs. And every backlog push borrows people from somewhere. Projects on schedule is usually where the loan shows.
A new portal doesn't settle those trade-offs on its own. GAO found that digital improvements made up 67 to 71% of the actions planned by federal High Impact Service Providers in each year from 2023 to 2025, and recommended that OMB set customer experience goals with quantitative targets and time frames. The same applies to a state agency's portal: it earns its budget when online share, first-time-right and the backlog all move on the same board.
So here's a test for any licensing division. Pick one open position and ask what it's costing in backlog days and minutes on hold. If the board can't answer, that seat looks like a saving in the next budget, and in a year when 14 states proposed cutting vacant positions, savings are what get cut.
Building it in SimpleKPI
Build this dashboard yourself with SimpleKPI's KPI dashboards, or start from the Library KPIs template. The mechanics are the same whatever you track:
- KPIs: each one gets a target, an entry frequency and tags, so you can report by team, site, product or whatever you split the numbers by.
- Data: type numbers in, import a spreadsheet, or use an integration or the API. The dashboard updates as each number lands, so daily and monthly KPIs sit side by side.
- Reports: scheduled KPI reports email the numbers to people who never log in.
- AI: Kip, the built-in AI, suggests targets and flags indicative benchmarks (AI estimates, not peer data) once some history builds up.
- Price: $79 a month for unlimited users and KPIs, as one flat price, with a 14-day free trial and no card needed.