Hiring Your First Technician

When to hire: the signals and the math

The first hire is the scariest check a founder writes, so most write it too late – after months of 60-hour weeks and slipping response times. Watch for the capacity signals: you are consistently more than 100% busy with billable work, you're personally absorbing 20–30 hours a week of steady overflow, tickets are aging, and you haven't taken a real day off in months. Sick days and vacations are the honest tell – a solo MSP has a bus factor of one, and clients eventually notice.

Then check the signals against the math:

• Labor revenue trigger: the common threshold is roughly $100k–125k in labor revenue – enough work to keep a tech busy from day one.
• Revenue per employee: the industry average benchmark sits around $142k per employee. If a hire drops you well below that, the hire dilutes profit rather than buying growth.
• The 3x rule: a technician should support roughly 3x their fully loaded cost in revenue – one third wages, one third overhead, one third profit. Fully loaded cost runs about 1.25–1.4x salary once taxes, benefits, and seat licenses are counted. Typical example: a $70k salary is ~$95k loaded and needs ~$210k+ of supported revenue.

If the MRR isn't there yet, fix pricing or sales before headcount – see KPIs and benchmarks for the numbers that should be true first.

Who to hire first

The majority path: a mid-level (L2-ish) technician. They're revenue-producing from early on, need minimal handholding, can face clients alone, and – critically – cover you on vacation and sickness, the real bottleneck for solo founders. A green L1 saves salary but consumes the founder time you're trying to buy back, because you become the trainer and the escalation path for everything.

The contrarian pick: an admin assistant. Billing, scheduling, invoicing, and vendor wrangling typically eat 10–15 founder hours a week, and an admin costs far less than a tech. If your calendar is drowning in admin rather than tickets, this is the better first hire. The practical synthesis most shops land on: tech first once billable work overflows, admin quickly after.

Sales last. Founder-led sales is the rule until roughly $1M+ revenue; first sales hires at small MSPs fail more often than not. A dispatcher role emerges later, around 3–4 techs.

What you'll pay (US, 2025–26)

Multiply by ~1.25–1.4x for fully loaded cost. Budget honestly: a "cheap" hire who can't work unsupervised costs more in founder hours than the salary gap.

The prerequisite: documentation and process

You cannot delegate what lives only in your head. Before the start date:

• Client documentation to your documentation standards: credentials in the vault, network diagrams, vendor contacts, backup configs. Documentation is what lets tech #2 onboard without tribal knowledge.
• Runbooks and SOPs for your ten most common ticket types – even one-page versions.
• A working ticket process: every request in a ticket, time logged on every ticket, statuses that mean something. Your new tech will copy whatever discipline they see, including the bad habits.
• A defined tool stack with the new tech's seats budgeted.

If none of this exists, spend a month building it before you post the job – it converts the hire from "apprentice to the founder" into "operator of a system."

Writing the job post

Small shops can't outbid corporate IT on salary, so sell what you actually have: variety (real networks, real servers, real security work – not one queue slice), direct client impact, proximity to the owner, and room to grow into a senior role as the company does. Be specific: name your stack, describe a real week, state the salary range. Skip the unicorn wish list – a requirements section demanding five certifications and seven years of experience for L2 money filters out exactly the practical, curious generalists a small MSP needs. Say plainly that it's a small company: some candidates will self-select out, and that's the filter working.

Interviewing: process over trivia

Certifications and memorized facts predict little. You're hiring for two things – troubleshooting process and customer communication:

• Walk me through it. Give a realistic scenario ("user can't reach a network share; nothing changed") and evaluate the method: do they gather symptoms, form hypotheses, test cheapest-first, and narrow scope – or guess randomly and reboot?
• The "I don't know" test. Ask something they likely can't answer. You want "I don't know, here's how I'd find out" – not bluffing. Bluffers create outages.
• Translation test. Have them explain a technical failure to you playing a non-technical office manager who is angry. Clear, calm, jargon-free explanation is the skill your clients will judge the whole company on.
• Ticket hygiene. Ask how they'd document the fix. A tech who writes nothing down doesn't scale past themselves.

Onboarding: shadow, then supervised, then owned

Do not hand over the queue on day one – it's the fastest way to burn a good hire and scare clients. Phase it:

1. Shadow (roughly the first couple of weeks): they ride along on your tickets, read the documentation, learn the stack and clients. Access provisioned, nothing owned.
2. Supervised queue: they work tickets you select – routine, well-documented types first – and you review every resolution and ticket note before close. Feedback is daily and specific.
3. Owned queue: they own day-to-day tickets with defined escalation criteria (time-boxed: stuck 30 minutes, escalate – no ticket-hogging, no hero culture). You stay on escalations, projects, and client relationships.

Move between phases on demonstrated competence, not the calendar, and introduce them to clients personally so the handoff builds confidence instead of doubt.

Utilization without burnout

A tech has ~2,080 paid hours a year; at a realistic 65–70% utilization that's roughly 1,350–1,450 productive client-facing hours. Average shops run 55–60%, top performers 75–80% – but chasing beyond ~80% backfires, leaving no slack for documentation, training, automation, or the inevitable P1 spike. Capacity-wise, a well-tooled tech supports roughly 250–400 managed endpoints, with ~350 often cited as the gold standard. Expect ramp-up: months, not weeks, to full productivity – that's why you hire on the leading edge of the math, not after you're drowning.

Bottom line

Hire when the math clears – roughly $100k–125k of labor revenue, a path to 3x their loaded cost – and the capacity signals are chronic, not a bad month. Default to a capable L2 who buys back founder time and covers your absences, pay the going rate (~$55–75k as of 2025–26), and spend the pre-hire month on documentation and process so you're handing over a system, not a mess. Then onboard in phases and manage to sustainable utilization. The first hire done well is the template for every hire after it.