Setting Salaries as a Small Business: What to Pay

Priya spent three weeks trying to fill a bookkeeper role at her landscaping company. She posted the job at $42,000 a year because it sounded reasonable to her, then watched it sit for two weeks with a single applicant who never showed up to the interview. When she finally called a friend who owns a similar company across town, she learned the going rate for that role in her metro was closer to $52,000. She had not lowballed the candidate on purpose. She had simply never checked what the number should have been.

That gap runs in both directions. Owners who guess low waste weeks on a role nobody wants at that price, then either raise it anyway or settle for someone underqualified. Owners who guess high, often out of nervousness about losing a candidate, lock in a payroll cost that gets harder to sustain every month the business does not grow into it, and one that is awkward to walk back once an employee has already accepted.

Getting the number right the first time does two things at once. It puts your money where it actually competes for the people you want, and it keeps them from leaving eighteen months later because a recruiter called with an offer 20 percent higher. Recruiting, interviewing, and training a replacement typically costs half of that employee's annual salary or more, so the research you skip at the offer stage tends to come back as a much larger bill later.

Do not set a number from a gut feeling, a number a friend mentioned once, or what you paid the last person in the role three years ago. Build the number from current data, specific to the role, the location, and the experience level you actually need.

Start broad, then narrow. Glassdoor, LinkedIn Salary, and Levels.fyi collect self-reported salaries by title, location, and experience (Levels.fyi is most useful specifically for tech and engineering roles). The federal Bureau of Labor Statistics runs its own Occupational Employment and Wage Statistics program, which updates more slowly than the crowdsourced sites but is far more rigorous, and it breaks numbers down by metro area. Indeed and ZipRecruiter often show a salary range directly on competitors' job postings, which is one of the fastest ways to see what companies like yours are actually offering right now, not what a survey said last year. Industry associations, from state restaurant associations to regional manufacturing groups, frequently publish their own compensation surveys, and those are worth checking because they capture your specific niche in a way a general salary site cannot.

One title can describe very different jobs. A marketing manager at a two person startup might be running every channel alone, while a marketing manager at a fifty person company might be managing three specialists and a budget. Read a handful of real job postings for the role, not just the title, to make sure the range you are pulling actually describes the job you are hiring for.

Pull a range, not a single number: the 25th, 50th, and 75th percentile for the role in your market. A single average hides how much room there actually is, and it gives you nothing to negotiate against when a candidate asks for more.

Market data gives you a starting range, not a final answer. A handful of factors then push that range up or down for your specific situation, and skipping this step is how owners end up paying a Kansas City rate for a role they are hiring in San Francisco, or the reverse.

Once you have a range, choose where inside it to land on purpose, rather than defaulting to whatever number a candidate first asks for.

Positioning below market, roughly the 25th to 40th percentile, only holds up if you are offering something else of real value: meaningful equity, unusually flexible working conditions, or a mission or brand candidates specifically want to be part of. It fails quietly if you use it as a way to avoid budgeting properly, because candidates who take a below market offer without a real reason tend to keep looking even after they start.

Positioning at market, roughly the 50th to 65th percentile, is the right default for most roles most of the time. It is competitive enough to attract solid candidates without overpaying for a role where you are not fighting off three other offers.

Positioning above market, the 75th percentile or higher, earns its cost when the role is genuinely critical, meaning a vacancy or a bad hire would be expensive, or when you are competing head to head against companies with more brand recognition than yours. Paying above market without one of those reasons is just an unexamined cost that compounds every payroll cycle.

Renata hired her first salesperson for her twelve person software company at a flat $50,000 salary with no commission, worried that a commission structure would feel complicated to set up. Six months later her best rep was fielding recruiter calls every week, because every other software company she competed with paid a base plus commission, and the total pay for hitting quota elsewhere ran about 30 percent higher than what Renata offered. Renata restructured to a 50 percent base, 50 percent at risk split tied to quota attainment, and kept the rep another two years.

For sales and business development roles, a base plus commission structure of roughly 50 percent base and 50 percent at risk is the norm, and deviating from it without a strong reason makes it harder to recruit experienced salespeople who already know what the market pays.

For non sales roles, performance bonuses typically run 5 to 15 percent of base salary for hitting clearly defined targets: a specific revenue number, a project shipped on time, a customer retention rate.

Avoid vague bonus promises. Telling an employee there might be a bonus if the company does well is demotivating rather than motivating, because it gives them nothing concrete to work toward and no way to know if they are on track. Tie any bonus to a specific, measurable outcome the employee can see themselves influencing day to day, and write it down before the year starts, not after.