The Coverage Factor: Why 4.2 People Don't Actually Cover One Post

The coverage factor is a multiplier. You take the number of people a post needs on paper, multiply it by the factor, and get the number of people you need on the roster. If a post needs 4.2 people and your factor is 1.156, you need 4.85, which means five. If you have been staffing to the raw number and feel short every week, this is the number you were missing, and it is specific to your team, which is why borrowing someone else's rule of thumb does not work.

The 24/7 coverage guide explains why 4.2 rounds up to five. This post is the calculation behind that jump: how to work out your own factor from last year's records, what to do when you do not have a year of records, and how to check your result against federal absence data.

What is a coverage factor, exactly?

A coverage factor is scheduled shifts divided by shifts actually available. Every person on your roster is scheduled for some number of shifts a year. Some of those shifts get lost to vacation, sickness, holidays, training and everything else that takes a person off the floor. What is left is the availability rate. The factor is its inverse.

• Availability rate = (scheduled shifts − shifts lost) ÷ scheduled shifts
• Coverage factor = 1 ÷ availability rate
• Roster need = raw headcount × coverage factor

You will also see this called a relief factor or a shift relief factor. It is the same idea. The name matters less than the discipline of deriving it from your own data instead of assuming it.

Two things about the formula are worth fixing in your head now. First, it works on shifts, or better, on hours, not on days. A person on three 12-hour shifts a week who takes a week of vacation loses three shifts, not five days. Second, every category of lost time sits in the same denominator, so you add the lost shifts together first and divide once. You do not multiply a vacation factor by a sick factor by a holiday factor. That double counts.

What data do you need to derive your own factor?

You need one year of records for the people who staffed the post, and four numbers from them.

1. Scheduled shifts per person. A standard five-shift week gives 260 a year. Count what you actually scheduled, not what the contract says.
2. Shifts lost to planned leave: vacation, paid time off and public holidays that fell on a scheduled shift.
3. Shifts lost to unplanned absence: sickness, family emergencies, jury duty, no-shows.
4. Shifts lost to everything else that pulls a person off the floor: required training, long leaves, and any shift where the person was on the roster but not covering the post.

Your time-off records and your timesheets hold all of this. If your team requests leave through a scheduling tool, the approved requests are the planned half. The unplanned half usually lives in timesheet exceptions or in a manager's notes, which is where the data gets patchy. Count it anyway. The rest of the calculation is only as good as this step.

Keep the categories separate even though the formula adds them. When the factor comes out higher than you expected, the split tells you which lever to pull: a policy decision about vacation, or an operational problem with unplanned absence.

How do you calculate the factor? A worked example

Take a five-person team that covers one post around the clock on eight-hour shifts. The raw need is 168 hours a week divided by a 40-hour week, which is 4.2 people. Each person is scheduled for 260 shifts, so the team was scheduled for 1,300 shifts last year. This is an illustration with round numbers, not a customer's data.

Here is what the year's records show:

The arithmetic runs in four steps.

1. Shifts available: 1,300 − 175 = 1,125.
2. Availability rate: 1,125 ÷ 1,300 = 0.865.
3. Coverage factor: 1,300 ÷ 1,125 = 1.156.
4. Roster need: 4.2 × 1.156 = 4.85, which rounds up to 5.

So five people is right, but only just. In an average week this team has about 173 hours of available labor against 168 needed, a cushion of roughly five hours. That is the number to sit with. Five people looks comfortable and is not.

The 24/7 guide walks through a lighter year of 18 days missed and lands at 4.5. This example has 35 shifts lost per person because it counts every category, vacation and holidays included. Neither is wrong. They are different teams with different leave policies, which is the whole argument for deriving your own.

How does a single long leave change the answer?

A coverage factor built on a normal year assumes absences are small and spread out. One long leave breaks that assumption, and it is the most common reason a team that did the math still ends up short.

Under the Family and Medical Leave Act, eligible employees can take up to 12 weeks of unpaid, job-protected leave in a year. To be eligible, an employee must have worked for the employer at least 12 months and at least 1,250 hours over the past 12 months, at a location with 50 or more employees within 75 miles. Check whether your headcount and locations put you over those thresholds.

Run the example again with one person taking the full 12 weeks. At five shifts a week that is 60 shifts lost.

• Shifts lost: 175 + 60 = 235
• Shifts available: 1,300 − 235 = 1,065
• Availability rate: 1,065 ÷ 1,300 = 0.819
• Coverage factor: 1,300 ÷ 1,065 = 1.221
• Roster need: 4.2 × 1.221 = 5.13

One leave moves the need from 4.85 to 5.13. Five people is now short, and the gap shows up as overtime or an uncovered shift for twelve weeks straight. The DOL overtime fact sheet confirms what that overtime costs: covered employees must receive pay at not less than one and a half times their regular rate for hours worked over 40 in a workweek. The overtime calculation guide shows how to price it.

You cannot predict which year has the long leave. What you can do is decide in advance how you will handle one: carry a sixth person, keep a trained float, or accept a planned stretch of overtime and budget for it. The point is to choose, not discover.

What do federal absence numbers say about typical ranges?

Use federal data as a sanity check on your own, not as a substitute. The Bureau of Labor Statistics publishes absence rates from the Current Population Survey in its absences from work table, broken down by occupation and industry. The figures below are 2025 annual averages for full-time wage and salary workers, as of November 2026.

Two columns matter. The absence rate is the share of workers who were absent in a given week. The lost worktime rate is hours absent as a percent of hours usually worked, which is the closest match to the shifts-lost share in your own records.

Read this table carefully, because it measures less than you might expect. BLS defines an absence as a full-time worker working fewer than 35 hours in the reference week for one of these reasons: own illness, injury or medical problems, child care problems, other family or personal obligations, civic or military duty, and maternity or paternity leave. It explicitly excludes vacation, personal days and holidays. The 2025 annual figures are also 11-month averages, because October 2025 data was not collected during the federal government shutdown.

So the table is a benchmark for unplanned absence only. The last column shows why it cannot be your whole factor: unplanned absence alone adds 1.5 to 2.2 percent, which is a factor barely above 1.01. The rest of the 1.156 in the worked example comes from planned leave, holidays and training that the survey leaves out on purpose.

That makes the table useful in a specific way. In the worked example, sickness plus other unplanned absence is 30 of 1,300 shifts, or 2.3 percent. Healthcare support workers lost 2.2 percent of their usual hours, so the example team sits a touch above its benchmark. That is not alarming. It is a prompt to look at whether the absences are concentrated in one or two people or in particular shifts, because an average that high usually has a cause.

How much planned leave should you assume?

Planned leave is the larger share of the factor, and it is a decision you made, not a statistic. Federal law does not require it. The Department of Labor states that the Fair Labor Standards Act does not require payment for time not worked, such as vacations, sick leave or federal or other holidays, and that these benefits are matters of agreement between an employer and an employee.

Access is common, though not universal. In the BLS Employee Benefits release for March 2026, 80 percent of private industry workers had access to paid vacation, 81 percent to paid holidays and 81 percent to paid sick leave. In health care and social assistance the figures were 82, 84 and 87 percent. The release text puts paid sick leave access as low as 58 percent in leisure and hospitality. Those are shares of workers with access, not average days taken, so they tell you how likely it is that your staff expects paid time off, not how many shifts to subtract.

For the number of days, your own policy is the better source. If you grant 15 vacation days and 6 paid holidays, a full-time employee will use close to all of them. Your time off policy and sick leave policy set the planned side of the factor directly. Changing the policy changes the headcount. A team that adds three vacation days per person has just raised its coverage factor, whether anyone did the math or not.

What if you do not have a year of history?

Build the factor from policy plus a benchmark. You will be less precise, and that is fine as long as you replace the estimate with real data as it accumulates.

Take a new post staffed on eight-hour shifts, five a week, 260 scheduled shifts per person. Your policy grants 15 vacation days and 6 holidays, and you expect about 3 training shifts.

1. Planned leave: 15 + 6 + 3 = 24 shifts, or 9.2 percent of 260.
2. Unplanned absence: use the BLS lost worktime rate for your industry. For health care and social assistance that is 2.1 percent.
3. Total lost: 9.2% + 2.1% = 11.3%, so availability is 0.887.
4. Coverage factor: 1 ÷ 0.887 = 1.128.
5. Roster need: 4.2 × 1.128 = 4.74, which rounds up to 5.

Adding the percentages here is a simplification, because the BLS rate is measured against hours usually worked, not shifts. At this level of precision the difference is small. Treat the answer as a starting point and tighten it.

After a quarter of real records, swap in your own unplanned rate. After a full year, swap in everything. Be careful with a partial year, because absence is seasonal. A winter quarter will overstate sickness and understate vacation, and a summer quarter does the reverse. If you have to annualize, annualize each category separately.

What does the factor not tell you?

A coverage factor answers one question: how many people does it take to keep a post covered, given how often people are not there. It does not answer several other things that move your real headcount.

• Vacancies and turnover. A seat that is empty for six weeks while you hire is lost coverage, but it is a different problem from absence. Track it separately and add it as its own allowance if your turnover is high.
• Part-time and mixed rosters. Calculate the factor in hours instead of shifts, and apply it to hours of demand. A mix of full-time and part-time staff (see the part-time and full-time mix guide) changes the math, because one person no longer covers one full-time slot.
• Clustering. The factor is an annual average. Absence is not spread evenly. Holidays, school breaks and flu season bunch it into particular weeks, and those are the weeks when you are short. If your roster is sized exactly to the average, the peaks are uncovered by definition.
• The shift pattern. A pattern with built-in rest days, like the ones in the shift pattern guide, changes your raw headcount before the factor is applied. Get the raw number right first, then apply the factor.

None of these invalidate the number. They tell you when to add to it.

What should you do with your number on Monday?

Pull last year's time-off and timesheet records for the team that staffs your busiest post, and sort the lost shifts into the categories above. That is an hour or two of work. Compute the availability rate and the factor, and multiply by your raw headcount.

Then compare the result to the people you have. If you are above it, you have slack, and the question is whether it is in the right weeks. If you are below it, the gap is already being paid for somewhere, usually in overtime, in managers covering shifts, or in the open shifts that nobody picks up.

Finally, check the unplanned share against the BLS benchmark for your industry. If it is well above, fix that before you hire to absorb it. Hiring a fifth person to cover chronic unplanned absence treats the symptom. Know your factor, know what is driving it, and revisit it once a year, or sooner if your leave policy changes.