Almost every piece of retention advice sorts businesses on one axis, size, and then draws a line through them. Which way the line runs depends on who is writing. Enterprise material implies that scale is what buys you the programs that keep people. Small-business material implies the opposite, that you are stuck competing against benefits you cannot match. Both are describing a slope.
The Bureau of Labor Statistics publishes the figure that settles it, broken out by establishment size, every month. Here is the quits rate for the private sector in July 2026, counted as quits during the month as a percent of employment, in Table 7 of the Job Openings and Labor Turnover release published September 1, 2026.
- 1 to 9 employees: 1.2 percent
- 10 to 49 employees: 2.2 percent
- 50 to 249 employees: 2.7 percent
- 250 to 999 employees: 2.2 percent
- 1,000 to 4,999 employees: 1.6 percent
- 5,000 or more employees: 1.0 percent
That is not a slope. It is a hump. The smallest employers and the largest ones sit next to each other at the bottom of the table, 1.2 and 1.0, and the worst quits rate in the private sector belongs to the band in the middle. Against a total private rate of 2.1 percent, a nine-person business is not losing people faster than average. On this month’s figures it is losing them at a little over half the average rate.
Which leaves a real question, because the pain is not imaginary. If the quitting is not unusual, what is?
The rate in that table that actually is high
Stay on the same page and look at job openings. For private establishments with 1 to 9 employees the job openings rate in July 2026 was 5.7 percent, the highest of any size class in Table 7, against 4.5 percent for the total private sector and 3.8 percent for the 50-to-249 band that has the worst quits problem. Then look at hires: 3.0 percent for the 1-to-9 band, the second lowest figure in that column, beaten only by the 5,000-plus giants at 1.9 percent.
Read the three lines together and they describe one business. Lowest quits. Highest share of roles standing open. Near-lowest rate of actually filling them. That is not a company with a leaving problem. It is a company with a refilling problem.
It also explains why the experience does not match the statistic. A rate is an average across a lot of businesses. Inside one business with nine people, a single resignation is eleven percent of the workforce and there is nobody to slide across from somewhere else. What decides the cost is not how often it happens, which the table says is rarely. It is how long the gap stays open afterwards, and the openings rate says it stays open a while. That quietly makes the speed of your hiring a retention question, because the people covering the gap are the ones who get tired. It is the argument for building the hiring process backward from the start date before you need it rather than during the scramble.
Where the curve turns
This next part is a reading rather than a finding, and I would rather flag that than smuggle it past you.
Step from the 1-to-9 band up to the 10-to-49 band and the quits rate nearly doubles, 1.2 to 2.2, then climbs again to 2.7 in the band above that. One month is one month, and these are preliminary figures, so nobody should build a plan on a single print. But the shape has a mechanism behind it, and the mechanism is not mysterious.
Whatever is holding people at nine employees is mostly you, personally, being able to see everyone. You notice when somebody goes quiet. You hear about the difficult client long before it turns into a resignation. None of that is a program and none of it is written down, which is exactly why it can stop without an announcement somewhere past the tenth hire. Nobody schedules the replacement of a system that was never installed.
So the honest use of a low quits rate is not reassurance. It is a deadline. The work below is worth doing in the year before you need it, which is the year you will least want to.
Three moments you already have
Retention programs assume a calendar a small business does not run: engagement surveys, quarterly cycles, somebody whose actual job this is. What a small business has instead is moments. These three already exist in your year, and each one is a point where a decision gets made about you whether or not you take part in it.
The week after the push ends
When a big project finally lands, most owners are already three days into the next one. That gap is where somebody privately decides whether the last quarter was an exception or an accurate description of the job. The useful question is narrow enough to answer: which part of the last three months would you not want to repeat. “How are things going” returns nothing, because it asks a person to volunteer a complaint with no container to put it in.
It is worth knowing what people say afterwards, when researchers ask them properly. Among US workers who quit a job in 2021, 63 percent said low pay was a reason, 63 percent said no opportunities for advancement, and 57 percent said feeling disrespected at work, in a Pew Research Center survey of 6,627 US adults fielded February 7 to 13, 2022. Two of those three describe things that could have been said out loud in time and were not. Write down the answer you get and what you decided to do about it, or the conversation did not really happen.
The moment you put a pay range in writing
Several states now require compensation to appear in the job posting itself, which means the number you offer a new hire becomes visible to the people already doing that work. Which states, and the headcount thresholds that catch a small employer, are in our piece on writing an employee value proposition on a small hiring budget. The retention half of it is simpler than the legal half: before the posting goes live, set the range beside what every current person in that band is actually paid, and decide what you are doing about anyone sitting below the bottom of it. You are going to have that conversation either way. The only variable is whether you open it or they do.
The sentence about what comes next
In a nine-person business, “no opportunities for advancement” almost never means there was no promotion to give. It means nobody ever said what the next thing was. That can be a single sentence, and the test of it is whether the person could check it in a year: not “room to grow” but which work moves to them, by when, and what has to be true first. Writing that down is most of what a development plan at this size actually is, and the free version is the one you say out loud and then send in an email so it exists.
What it costs to make time off real
Paid leave is the one column where a small employer is close to a large one, because leave is granted rather than purchased. It only works as a benefit if people take it, though, and on a small team they often do not, for a reason that has little to do with culture. Nobody can see who else is out, so every request feels like it might be the one that breaks the week. Making the calendar visible is a cheaper fix than it sounds.
Timetastic publishes three tiers on its pricing page: Business at $1.50 per user per month, Pro at $2.50, and Rota and Time Clock at $5.50, with a 30-day free trial and no contract. For nine people the entry tier is $13.50 a month. Before adding it, check what payroll already does, because Gusto lists Simple at $49 a month plus $6 per person and names basic PTO policies and holiday pay among that tier’s features, with Plus at $80 plus $12 per person and Premium at $180 plus $22 per person, on its own pricing page. The common waste here is buying a time-off tool to sit beside a payroll product that already tracks time off.
The free version is a shared calendar with one entry per day away, visible to the whole team. The mechanism that matters is visibility, not the software. Pay for a tool when the spreadsheet starts producing arguments, and not before.
What none of this does
There is a tempting misreading of a 1.2 percent quits rate, which is that it represents slack. Run leaner, nobody leaves anyway. The same table refuses that reading. A business with the lowest quits rate in the private sector and the highest share of roles standing open is already running short-handed, and the people still there are absorbing work a larger company would have hired for.
That is worth being blunt about, because it points the opposite way from where AI conversations usually end up. The gap between those two figures is a case for filling the role and paying the people currently covering it, not a case for discovering how far a short team will stretch. An assistant is genuinely useful for the drafting: turning scrappy notes into a development sentence somebody can check, or a first pass at a leave policy. It cannot tell you whether the sentence is true, and a confident plan that commits you to something you have not decided to fund is worse than no plan. Read anything it gives you as an editor whose only job is deleting what you cannot back.
If the blank page is the obstacle, BusinessPrompter publishes a free prompt chain, Reduce Employee Turnover and Burnout, which works through the diagnosis before it asks you to build anything. The wider library is at BusinessPrompter.com. The same editing rule applies to its output as to any other draft.
Where to start
Open Table 7, find the size band you are actually in, and write its quits rate next to the number of people who left you in the last twelve months. Then write the job openings rate for the same band underneath. Most owners find the first pair less alarming than they expected, and the second number the one that describes their year.
Questions owners ask about this
Is turnover really lower at small businesses?
On the most recent figures, yes. In July 2026 the quits rate at private establishments with 1 to 9 employees was 1.2 percent, against 2.1 percent for the total private sector and 2.7 percent at establishments with 50 to 249 employees, per BLS Table 7 published September 1, 2026. Those are monthly rates and one month is not a trend, but the pattern across that table is a hump rather than a straight line by size.
What do people actually say when they quit?
Among US workers who quit a job in 2021, 63 percent gave low pay as a reason, 63 percent gave no opportunities for advancement, and 57 percent gave feeling disrespected at work, in a Pew Research Center survey of 6,627 US adults fielded February 7 to 13, 2022. Pay and advancement are both things a small employer can put in writing before anyone has to ask.
What is the cheapest retention change a small employer can make?
Making time off visible so people actually take it. The free version is a shared calendar with one entry per day away. If you want a tool, Timetastic lists Business at $1.50 per user per month with a 30-day free trial on its pricing page, but check payroll first: Gusto’s Simple plan at $49 a month plus $6 per person names basic PTO policies and holiday pay among its features, on its pricing page.
If my quits rate is already low, why does it feel so bad?
Because a rate is an average and a resignation is not. In a nine-person business one departure is eleven percent of the workforce. The figure that matches the feeling is in the same table: a 5.7 percent job openings rate for the 1-to-9 band, the highest of any size class, alongside a 3.0 percent hires rate, per BLS Table 7. The role stays open, and the people covering it carry the cost.
