On June 9, the NFIB put out its May 2026 Small Business Optimism survey, and it hit a record nobody was hoping for.
For the first time in the survey's 52-year history, owners ranked labor costs as their single most important problem. 14% of them, up 5 points in a single month. It even passed labor quality for the first time since 2013, which means owners are now more worried about what people cost than whether they can find good ones.
Here's the part that should make you stop and look twice. While owners are panicking about the cost of labor, they're hiring less, not more.
The numbers don't add up the way you'd expect
Unfilled job openings dropped to 29% of owners, down 5 points, the lowest since May 2020. The share planning to add jobs in the next three months fell to a net 9%, also the lowest since 2020. Capital spending plans hit 16%, the worst reading since March 2009.
So this isn't a labor shortage anymore. Owners aren't desperate for bodies. They're pulling back on every kind of fixed commitment they can.
And wages barely moved. Net pay raises ticked up a single point. If salaries aren't spiking, where's the record cost anxiety coming from?
It's coming from the structure, not the paycheck. In 2026, 19 states raised their minimum wage, and in several of them the overtime-exempt salary threshold is tied to that number, so it climbed too. NFIB's chief economist pointed straight at "costly new state mandates." The squeeze isn't the salary line. It's everything bolted around it.
What a full-time seat actually costs
When you hire someone, the number on the offer letter is the part you can see. It's also the smaller part.
The old SBA rule of thumb still holds: a full-time employee runs 1.25 to 1.4 times their base salary once you add it all up. Benefits alone are about 30% of total compensation. So a $100,000 hire is really a $125,000 to $140,000 commitment. The salary is roughly two-thirds of the real cost.
Now stack the risk on top. The Department of Labor pegs a bad hire at around 30% of first-year salary, and that's the floor. CareerBuilder puts the average loss near $17,000. For a two or three person shop, one wrong hire can wipe out a whole quarter's budget for tools and training.
That's the thing about a full-time seat. It's not just expensive. It's a big, lumpy, hard-to-reverse bet, and small teams feel every miss the hardest.
The "tech person" trap
The clearest place this shows up is the role most small businesses get wrong: the in-house tech person.
The math goes like this. A single IT generalist costs somewhere around $110,000 to $135,000 fully loaded, plus another $15,000 to $30,000 a year in tools and licenses. And here's the catch. One person is one skill set and one calendar.
The job is cloud, security, compliance, email, backups, networking, and the help desk. No single hire is genuinely good at all of those. They're underused on a quiet Tuesday and underwater the day something breaks. When they're out sick, your coverage is zero. When they leave, all the knowledge walks out with them.
That's what I mean by buying a full-time seat for part-time work. You're paying permanent, fixed cost for a function that's specialized and spiky. The work is real. The full-time seat is the part that doesn't fit.
The fix is noticing, not hiring cheaper
The instinct when labor costs spike is to look for a cheaper hire. That's the wrong move. The win isn't a discount on the seat. It's spotting which roles were never a hire to begin with.
There's a simple test: core versus context, continuous versus spiky.
If the work is core to your business and runs every single day, keep it in-house. A developer building the product your company sells should be on your team. That's not a fractional role, and pretending otherwise is how things go sideways.
But infrastructure, security, the tech that has to work but isn't what you sell? That's context. It's specialized, it spikes, and it does not need a permanent body in a chair. You need the capability, not the headcount.
This is already the direction the market's moving. Fractional and interim placements are up 310% since 2020. More than half of US small businesses now outsource at least one core function. None of this is a prediction. It's owners doing the math you just did.
How we actually run this
I'll be straight about why this isn't theory for us. Kief Studio is two people. Me and my wife Meelie. That's the entire company.
Between us, the combined skill set covers what would normally take a 10 to 14 person team, roughly $1.15M to $1.7M a year in salaries for comparable coverage. We don't carry that headcount. We built a system we call LTFI, our own AI tooling, to automate the roles we'd otherwise have had to hire for, and we augmented ourselves to do the work of several disciplines.
So when a small business brings us in, they're not renting one generalist and hoping. They get build, security, automation, and support as a function, with people who actually specialize in each, and none of the fixed overhead, mandates, or single-point-of-failure risk that comes with a permanent seat. One team for everything, on a cost you can turn up or down.
That's the whole reframe behind the NFIB record. The report doesn't say people are too expensive. It says the full-time employment structure is too expensive to staple onto work that's specialized or part-time. Keep your core, continuous work in-house. Stop buying permanent seats for the rest.
If you're staring at a role right now and can't tell if it's a real hire or a seat you don't actually need, that's exactly the conversation worth having. The first one is free, no commitment.
Subscribe at kief.studio for free, and you'll get the companion resource on sorting core-versus-context roles, plus the rest of our guides and tools.