214 decisions in ten working days. That is the count one owner of an 11-person residential service company got after writing down every question, approval, and judgment call that landed on them for two weeks. If you're trying to figure out how to build a management layer in a small business, don't start with an org chart or a job title. Start with a number like that one, because it tells you exactly what a manager needs to take off your plate, and in what order.
A note on the example before we begin. The company below is a composite. It's built from a pattern I've watched repeat in a lot of service businesses, and I've rounded the figures to keep the math readable. The method is exactly what I'd run on my own business, and you can run it on yours starting Monday.
The setup: 11 people, one decision-maker
The company has one owner, one office coordinator, and nine field technicians working in three crews of three. Revenue is healthy, and reviews are good. The owner is working 60-hour weeks anyway, and the crews say the same thing: "We're always waiting on a call back."
The structure is flat, which means everyone reports to the owner. That's 10 direct reports. With 11 people, a team has 55 possible one-to-one communication lines (the formula is n × (n - 1) ÷ 2). In a flat company, most cross-team questions don't travel between two employees. They route through the owner, so the owner becomes the switchboard for all 55.
Flat works fine at four or five people. Somewhere between 8 and 12 it stops working. There's nothing magic about those headcounts. The volume of daily judgment calls just outgrows one person's attention before it outgrows anything else.
If this sounds like your week, see how owners hand this off.
When does a flat small business need a management layer?
A small business needs a management layer when owner decisions, not labor or sales, become the constraint on how much work gets done. You'll see it before you measure it:
- Crews finish a task and stand around waiting for the next instruction.
- The same question gets asked by three different people in the same week.
- You're the only person who can approve a change order, a material run, or a schedule swap.
- Taking a half day off means your phone is fuller when you come back than if you'd stayed.
Symptoms don't tell you what to hand off, though. The two-week decision log does. For ten working days, the owner kept a running note on their phone. Every time someone asked for a decision, they logged the time, who asked, and a two-word category. No analysis, just tally marks. (If the interruptions themselves are what's burning you out, how to stop answering every question from your team covers the triage side of this.)
Which decisions should a management layer take first?
Here's what the log showed after ten working days:
| Decision category | Count (10 days) | Per day | Who could own it? |
|---|---|---|---|
| Scheduling and dispatch changes | 71 | 7.1 | Office coordinator |
| Job-site judgment calls ("do we replace or repair?") | 58 | 5.8 | Field lead |
| Materials and supply purchases | 31 | 3.1 | Field lead, within a limit |
| Customer complaints and callbacks | 23 | 2.3 | Split: field lead first response, owner on refunds |
| Time off, attendance, people issues | 17 | 1.7 | Owner for now |
| Pricing and scope changes | 14 | 1.4 | Owner |
| Total | 214 | 21.4 |
Now put a time cost on it. A decision that takes 90 seconds to answer still costs about six minutes once you add the context switch: stopping what you were doing, finding the job details, answering, and getting back to where you were. So:
- Owner time: 21.4 decisions × 6 minutes = about 128 minutes a day, or roughly 10.7 hours a week spent as a switchboard.
- Crew wait time: the 58 job-site calls had a median wait of about 12 minutes before the owner called back, with two techs idle each time. 58 × 12 minutes × 2 people = 1,392 labor-minutes, or about 23 hours of paid, unbillable time in two weeks.
That second number is the one that gets an owner's attention. Twenty-three crew hours every two weeks disappear into waiting for you. Nobody logs that as a cost, and it's still a cost.
The table also answers the "what first" question. You don't transfer authority by title. You transfer it by category, starting with the highest-volume, lowest-risk decisions. Scheduling (71) and job-site calls (58) are 60 percent of the load. Pricing (14) and people issues (17) are only 14 percent, and they're the ones with the most downside if a new manager gets them wrong. So the first two move now, and the last two stay with you.
Who can supervise when you build the first management layer?
This is where most owners make the expensive mistake of promoting their best technician. According to Gallup's research on why great managers are rare, only about one in ten people have high talent to manage, and companies pick the wrong person for manager roles 82% of the time. The best producer is often a poor fit. They're fast because they don't need to explain things, and managing is mostly explaining things.
In the example company, the fastest tech was the obvious pick. Instead, the owner went back to the decision log and asked a different question: when I didn't answer within 15 minutes, who did the crews call next? One name came up again and again. It was a tech with six years in, not the fastest, but the one who:
- Already got texts from other crews asking "what would you do here?"
- Wrote notes on job tickets without being asked.
- Had once told a newer tech, respectfully, that their cleanup wasn't good enough, and the newer tech fixed it.
- Gave credit to the crew when a customer complimented a job.
Those four traits are the screen. The team already sees this person as the lead, they document things, they'll have the uncomfortable conversation, and they don't need the spotlight. Skill at the craft is the floor, not the tiebreaker.
Before making it official, run a two-week trial. Call it an "acting lead" role. You're still available, but when a job-site question comes to you, you answer, "What does Dana think?" and route it back. By day ten you'll know whether the crews accept the call and whether the acting lead makes it without freezing.
What management authority actually transfers to the new lead?
A title with no authority behind it is how new managers fail. The crews figure out within a week that the lead's answer can be overruled, and they go straight back to you. The fix is a written decision-rights sheet with four levels:
| Level | What it means | Example categories (field lead) |
|---|---|---|
| Decide | Makes the call and doesn't report it | Crew assignments within the day, repair-vs-replace on standard jobs |
| Decide and tell | Makes the call and reports it in the daily or weekly summary | Material purchases under the per-job limit, first response to a callback |
| Recommend | Brings a written recommendation, and the owner decides | Discipline, raises, whether to take on an unusual job |
| Owner only | Stays with the owner | Pricing, scope changes, refunds, hiring and firing |
Two rules make the sheet work. First, every "Decide and tell" item has a number attached. The purchase limit gets written as an actual figure you picked, not "reasonable purchases." Vague limits get escalated anyway. Second, you don't overrule a "Decide" call in front of the team. If the lead got it wrong, you talk about it privately and adjust the sheet. Overruling in public undoes the whole layer in one afternoon.
This sheet sits on top of an accountability chart, not in place of one. If you haven't drawn one yet, how to build an accountability chart for a small business covers who owns each function. The decision-rights sheet then defines how much authority each seat carries.
The legal lines a new supervisor crosses
Promoting someone changes more than the org chart. There are three things to check before you announce it:
- Overtime status. Many owners assume a manager should go on salary. Under the Department of Labor's executive exemption rules (Fact Sheet #17B), a manager has to meet a salary threshold, have management as the primary duty, and regularly direct two or more full-time employees or the equivalent. A field lead who still spends most of the day doing the work is usually a "working supervisor" under 29 CFR 541.106 and stays hourly with overtime. The common setup is a lead differential on the hourly rate, not a switch to salary.
- Harassment training. Some states attach training requirements to the supervisor title. In California, employers with five or more employees must give each new supervisor two hours of harassment prevention training within six months of the promotion. The California Civil Rights Department publishes the requirements and a free course. Check your own state's rules.
- Liability. Once a lead can hire, fire, or reassign people, courts treat them as a supervisor for harassment liability purposes. The Supreme Court drew that line in Vance v. Ball State (2013). That's one more reason "hiring and firing" stays in the Owner-only or Recommend column for the first year.
How do you support a first-time manager in a small business for the first 90 days?
New managers don't fail because they're bad at the job. They fail because they were handed the job and left alone with it. In the example company, the owner ran a three-phase ramp:
- Days 1 to 30: decide, report daily. A 10-minute end-of-day check-in covers what calls the lead made, which ones felt shaky, and what they escalated. Expect week one to look worse. Our lead escalated roughly 40 percent of job-site calls in the first week. By week four it was under 15 percent.
- Days 31 to 60: decide, report weekly. The daily check-in becomes a 30-minute weekly one-on-one. The lead brings a short list of the decisions they made and the ones they weren't sure about. You coach on the unsure ones.
- Days 61 to 90: decide, report exceptions. The one-on-one stays weekly, but the agenda shifts from "what did you decide" to "what's coming up that you'll need from me." At day 90 you review the decision-rights sheet and move one or two categories up a level.
Alongside the ramp, do three more things. Make the change visible: announce it to the whole team, give the lead a title and pay change, and tell customers who the field point of contact is. Train the lead on the people part, not the technical part. How to give correction, how to run a two-minute morning huddle, and how to document a problem are the skills they've never been taught. And when a tech goes around the lead to you, send them back every time with the same line: "What did Dana say?" If that's hard for you, the patterns in the delegation roadmap for service business owners will feel familiar. The redirect is most of the job.
The after picture: running the numbers again
At day 90 the owner kept the same phone log for another ten working days. Here's what moved:
- Scheduling and dispatch (71): fully with the office coordinator, who was already doing half of it without the authority to finish it.
- Job-site calls (58) and materials (31): with the field lead, inside the written limits.
- Complaints (23): the field lead handled first response and resolution on about 15, and the owner kept the 8 involving refunds or angry customers.
That's 175 of 214 decisions transferred. The owner still handles about 39 over ten days, or 3.9 a day, down from 21.4, which is an 82 percent reduction. Using the same six-minute cost:
- Before: 128 minutes a day. After: about 23 minutes a day. That's roughly 105 minutes a day, or about 8.75 hours a week, back.
- Subtract the new management overhead: two 30-minute weekly one-on-ones (coordinator and field lead) cost 1 hour a week. The net is about 7.75 hours a week recovered after day 90, and about 6.9 hours during the first month while the daily check-ins were still running.
- Crew wait time dropped because the lead now answers from a truck on site, usually within two or three minutes. Most of those 23 idle hours every two weeks went back to billable work.
Notice what didn't happen. The owner didn't get their week down to 30 hours. They got about a full workday back and spent it on the things still in the Owner-only column: pricing, hiring, and the relationships that bring in work. That's the point of building the layer. It moves you up to the decisions only an owner can make. If you want to lock those hours in, set up a weekly operating rhythm for them. Otherwise the reclaimed time fills back up with small stuff.
Where this example stops working
One field lead supervising eight techs is near the ceiling for someone who still does production work. At 14 or 15 people, the next move is two crew leads reporting to one operations manager, and you run the same log-sort-transfer process one level down. The layer doesn't get built once. It gets rebuilt each time the headcount outgrows the one you have. Good books, a clear job-tracking system, and sensible automation all make each new layer easier to hand over, because the new manager is inheriting a process instead of your memory. The people side of this, meaning pay bands, reviews, and career paths, is covered in people systems for a small business.
Questions owners ask about building a management layer
Should I promote my best technician to manager?
Not by default. Look for the person the team already calls when you're unavailable, who writes things down, and who will have an uncomfortable conversation. Craft skill is the minimum requirement. It shouldn't decide the promotion.
Do I have to put my new supervisor on salary?
Usually not, and often you shouldn't. A lead who still spends most of the day doing field work generally doesn't qualify for the FLSA executive exemption and stays hourly with overtime. A lead differential on their hourly rate is the common setup. Confirm the details with an employment attorney or your payroll provider.
What if my team keeps coming to me instead of the new manager?
Send them back every time with "What did your lead say?" Also check that the decision-rights sheet actually gave the lead authority over that category. Teams go around a manager when they've learned the manager can be overruled.
How many people can one supervisor manage in a service business?
A working lead who still does production work can usually handle about six to eight people. A full-time manager with no production duties can handle more. When the lead's own escalations start climbing again, the span is too wide and it's time to split it.
Should I hire an outside manager instead of promoting from within?
Promote from within when someone already has the team's trust. Hire from outside when nobody passes the trial, or when the role is mostly process and systems rather than leading crews. Either way, the decision-rights sheet and the 90-day ramp still apply.
Frequently asked questions
Should I promote my best technician to manager?
Not by default. Look for the person the team already calls when you're unavailable, who writes things down, and who will have an uncomfortable conversation. Craft skill is the minimum requirement. It shouldn't decide the promotion.
Do I have to put my new supervisor on salary?
Usually not. A lead who still spends most of the day doing field work generally doesn't qualify for the FLSA executive exemption and stays hourly with overtime. A lead differential on their hourly rate is the common setup. Confirm the details with an employment attorney or your payroll provider.
What if my team keeps coming to me instead of the new manager?
Send them back every time with "What did your lead say?" Also check that the decision-rights sheet actually gave the lead authority over that category. Teams go around a manager when they've learned the manager can be overruled.
How many people can one supervisor manage in a service business?
A working lead who still does production work can usually handle about six to eight people. A full-time manager with no production duties can handle more. When the lead's own escalations start climbing again, the span is too wide and it's time to split it.
Should I hire an outside manager instead of promoting from within?
Promote from within when someone already has the team's trust. Hire from outside when nobody passes the trial, or when the role is mostly process and systems rather than leading crews. Either way, the decision-rights sheet and the 90-day ramp still apply.