The Short Answer and Why It Varies
Endpoints per technician is one of the most quoted ratios in the MSP world and one of the least comparable. A 2025 Acronis benchmark article calls 350 fully managed endpoints per technician the gold standard, and in the same breath notes that the figure varies with service complexity and the technology stack. That caveat is the whole story.
Two MSPs can report the same endpoints per tech and be running completely different businesses. One supports a standardized fleet of laptops for professional offices with strong automation. The other supports manufacturing floors, line-of-business servers and a dozen different firewall brands. The first MSP is comfortable; the second is burning out its team. The useful question is not whether you hit a published number, but what your own team can carry while response times, backlog and morale hold steady. That number is yours to discover, and it moves as your stack, clients and automation change.
What the Benchmarks Actually Say
Kaseya's 2023 MSP Benchmark Survey, with 1,091 respondents, asked both technicians and executives how many endpoints technicians manage. The answers did not match. The most common answer from technicians was more than 750 endpoints, chosen by 26%. The most common answer from executives was 101 to 250 endpoints, chosen by 27%, and only 8% of executives put the figure above 750.
| Source | What It Reports | How to Read It |
|---|---|---|
| Acronis (2025) | 350 fully managed endpoints per technician as the gold standard | A target for a mature, standardized MSP, not an average |
| Kaseya survey, technicians (2023) | 26% said they manage more than 750 endpoints | Possibly counting every device in the RMM |
| Kaseya survey, executives (2023) | 27% said 101 to 250; only 8% said more than 750 | Possibly dividing by total technical headcount |
The gap between those two answers is a useful warning. Before you compare your ratio to anyone else's, decide exactly what you are counting on both sides of the division.
Define the Ratio Before You Compare It
Most disagreement about this number is really disagreement about definitions. Settle these before you track the metric month to month:
- Which endpoints count. Fully managed workstations and servers under a managed agreement, or every agent in your RMM, including printers, network gear and devices you only monitor?
- Whether servers count the same as laptops. A domain controller or a line-of-business server generates far more work than a standard laptop, and you can weight them.
- Which people count. Only service desk technicians, or also escalation engineers, project engineers, the service manager and the owner when they work tickets?
- Whether project hours are included. Onboarding and project work consume technician time without adding managed endpoints.
Pick a definition, write it down, and keep it. The trend in your own ratio over time is more useful than any external comparison, because it is measured the same way every month.
What Pushes Your Number Up or Down
The ratio is an outcome, not a lever. It rises when each endpoint needs less human time, and it falls when each endpoint needs more. The main drivers are listed below.
| Driver | Raises Endpoints per Tech | Lowers Endpoints per Tech |
|---|---|---|
| Stack standardization | One firewall, one backup product, one EDR across clients | A different vendor at every client |
| Client environment | Cloud-first offices on current hardware | On-premises servers, legacy apps, aging hardware |
| Automation | Scripted patching, onboarding and remediation | Manual fixes repeated ticket after ticket |
| Client selection | Clients who accept your standards | Clients who refuse upgrades or demand exceptions |
| Documentation | Current runbooks and client documentation | Knowledge that lives in one senior tech's head |
| Security posture | Strong baseline with MFA and least privilege | Frequent incidents and cleanup work |
Notice that most of these are decisions you make at the sales and onboarding stage. A client who joins without meeting your standards will cost you technician capacity for as long as they stay. The client onboarding checklist is where you can enforce those standards.
Measure Load, Not Just Devices
Endpoint count is a proxy for workload, and a weak one on its own. Pair it with tickets per endpoint per month and tickets closed per technician per day. If tickets per endpoint fall, each technician can carry more devices without working harder. If tickets per endpoint rise, adding endpoints will break the desk long before you reach any benchmark.
Watch the service quality signals at the same time. When a team is overloaded, first response time slips first, then aged ticket backlog grows, then employee turnover follows. A high endpoint ratio that comes with rising backlog is not efficiency; it is deferred cost.
- Pull a month of closed tickets from your PSA and count managed endpoints from your RMM.
- Divide tickets by endpoints to get tickets per endpoint per month.
- Divide endpoints by the technical staff who work tickets, using your written definition.
- Compare both numbers to last quarter, and check backlog and response times for the same period.
- Only raise your target ratio when tickets per endpoint are falling and service levels are holding.
Signs Your Team Is Past Its Limit
Benchmarks tell you what is possible for someone else. Your own team tells you what is possible for you, usually before any spreadsheet does. Technicians rarely say outright that they are overloaded; the signs show up in how the work gets done.
- Tickets are closed quickly but reopen, so ticket reopen rate climbs while closure counts look healthy.
- Proactive work stops. Patch reviews, backup checks and documentation updates slide because reactive tickets take every hour.
- The same senior technician is pulled into every escalation, and nobody else learns the hard problems.
- Onboarding a new client pushes existing clients' tickets back by days.
- People stop taking time off, or take it and return to a pile they cannot clear.
When two or more of these appear together, treat the ratio you are running at as your ceiling for now, whatever the benchmark says. Adding endpoints past that point does not raise efficiency; it moves cost into turnover, errors and client churn, which are harder to see on a monthly report and much more expensive to fix.
The opposite signal matters too. If backlog is near zero, technicians have time for projects and documentation, and response times are steady, the team probably has room. That is the moment to take on the next client or two before hiring, and to check whether the capacity comes from real efficiency gains or from a quiet month.
Using the Ratio to Plan Hiring and Pricing
Once you know your own sustainable ratio, it becomes a planning tool. Divide your managed endpoint count by that ratio and you have the technical headcount you need today. Add the endpoints in your sales pipeline and you can see when the next hire is due, which is the core of the staffing calculator.
The ratio also feeds pricing. If one technician can carry a known number of endpoints, you know the labor cost behind each one, and you can check that your all-in seat price covers it with room for margin. The managed services pricing calculator works through that link, and how MSPs price their services explains the models.
If your ratio is well below where you want it, resist the urge to fix it by not hiring. Fix the drivers first: standardize the stack, automate the repeat tickets, and tighten onboarding. The ratio will follow, and your team will still be there when it does.
Common Questions
Should servers count as more than one endpoint?
You can weight servers because they generate more work than a workstation. Whatever you choose, apply it consistently so your ratio can be compared month to month.
Do mobile phones count as endpoints?
Only if you actively manage them, for example through mobile device management with policies and support. Phones that just have email set up usually add little workload and are often excluded.
Does co-managed IT change the ratio?
Yes. In co-managed arrangements the client's internal staff handle some of the work, so your technicians can often cover more devices, but the ratio is not comparable to fully managed clients.
Is a very high ratio always a good sign?
No. A high ratio paired with growing backlog, slower responses or rising turnover usually means the team is overloaded rather than efficient.
How often should I recalculate the ratio?
Monthly is common. Recalculate after any large client onboarding or offboarding, since a single change can move the ratio noticeably in a small MSP.
Does outsourcing the help desk affect it?
It changes who does the work. If an outsourced desk handles first-line tickets, count only your own staff and track the outsourced cost separately so the ratio stays honest.