This case study is based on a real Fusion Computing engagement.
Client details have been anonymized, and certain technical details
have been generalized for privacy.
I got the call after the studio’s fourth production-hour outage in
two months. They were a 35-person Toronto design firm. Engagements like this one shape how we structure our wider Fusion Toronto managed IT engagements for growth-stage studios and agencies. They had signed
work, a hiring plan that would push them past 60 employees by year-end,
and one capable IT lead who was already the bottleneck.
The CEO said one sentence I still remember: I don’t think
we’re going to make it through this year if I don’t fix IT now.
He was right. What followed became a five-year co-managed engagement
that took the studio from 35 users to 205 users without IT ever becoming
the constraint that paused hiring or delivery.
SNAPSHOT · 35 TO 205 USERS · 2021 TO 2025
| Engagement | Co-managed IT, Toronto design studio |
| User scale | 35 to 205 users in under 4 years (about 6x) |
| Infrastructure | Ground-up rebuild: network, servers, MDM, security stack, racks |
| Onboarding | 4 to 8 hours per hire dropped to 45 minutes (about 85% reduction) |
| Help desk FCR | 62% to 91% first-contact resolution |
| Security incidents | 2 events in prior 12 months to zero across 3 years post-build |
| Cost per user month | Down 24% despite higher security coverage |
| Internal IT headcount | 1 internal lead, unchanged (co-managed, not replaced) |
| Engagement length | Active 5+ years and still running |
Source: Fusion Computing project records, Toronto design studio engagement, 2021 to 2025.
What I walked into at 35 users
Across the Fusion Computing engagements I run for Canadian SMBs in
the 30 to 100 user range, I see roughly the same four-problem stack
about 70% of the time. This studio was textbook.
The fleet was mixed Windows and Mac with no consistent management.
Some Macs were enrolled in Jamf. Most Windows laptops weren’t enrolled
in anything. There was no MDM baseline, so patching was a guess.
Onboarding took 4 to 8 hours per hire because every setup was
bespoke. Backup was fragmented, with different tools for different
departments, no tested restore, and no offsite copy of the production
NAS. MFA was on Microsoft 365 only. About half the SaaS tools that
actually held client data weren’t covered.
The CEO didn’t need me to find the problems. He needed someone to
tell him whether the architecture had any chance of carrying twice as
many people, then six times as many. The honest answer was no. Not
without a rebuild.
What had to scale and what couldn’t
I told the CEO four things had to change before headcount doubled,
not after. Large project files moved across the network all day, so the
network design and storage layout were no longer optional. The hardware
profile mattered: designers needed GPU, RAM, and storage performance
that commodity office laptops don’t deliver, and lifecycle planning
had to reflect that.
Deadlines were fixed, so downtime had real cost. And after-hours
work was real, so a 9-to-5 IT model was already behind the business.
The internal IT lead was capable. The job, at 35 users heading toward
100, was already too big for one person. That’s the gap a co-managed
model is built for, and it’s the call we made together: co-managed IT services
rather than full outsourcing.
What I built and how the co-managed split actually worked
I told the CEO upfront we weren’t going to replace his IT lead. He
was the one piece of this that already worked: he knew the team, he
knew the workflows, and he was the face of IT inside the business.
Replacing that proximity with a fully external model would’ve
created a different kind of gap. So Fusion Computing built a
co-managed structure with a clean division of responsibilities, and the
internal lead stayed exactly where the business needed him.
The split looked like this:
| Internal IT lead owned | Fusion Computing owned |
|---|---|
| L1 support, frontline troubleshooting, workstation readiness, hands-on local issues, the day-to-day relationship with users. | Monitoring, patching, security operations, escalation, vendor coordination, backup oversight, infrastructure lifecycle, strategic guidance. |
That split did two important things. First, it removed
single-person dependency on critical systems. Second, it made the
internal lead more effective: instead of drowning in routine
operational drag, he could focus on the business-facing work that
actually benefits from being in-house.
Fusion Computing also trained him to our operational standards:
how to document, how to handle tickets with consistency, how to follow
patching schedules, and how to escalate the right issues at the right
time.
Five years in, the model still works. That’s the point. Plenty of
IT setups look good in the first 90 days. Very few survive growth,
hardware refresh cycles, vendor changes, and shifting business demands.
This one did.
The three-layer stack we landed on
Once the operating model was clear, the technology choices fell out
of it. Fusion Computing standardized the environment around three
layers, each one chosen because it had to scale to 200 users without
re-architecture.
| Layer | What we deployed |
|---|---|
| Identity and device | Microsoft Entra ID for SSO and conditional access. Intune for Windows MDM. Jamf for Mac MDM. Device compliance enforced before any SaaS access. |
| Security stack | EDR on every endpoint. Microsoft Defender for Office 365 email security. MDR monitoring against the CIS Controls v8.1 baseline. MFA on every account that touched client data. |
| Network and platform | Fortinet firewall, structured cabling sized for the studio’s production workload, hybrid Azure plus on-prem servers. Performance-sensitive workloads stayed local. Backup, DR, and selected services moved to Azure. |
Fusion Computing’s project records show zero unplanned downtime
across the 4-month phased deployment.
Three years and four numbers later
The build phase was the easy part to plan. The proof is what happened
over the three years that followed. Here’s what changed in the studio’s
operating numbers:
| Metric | Before | After |
|---|---|---|
| User count | 35 | 205 (about 6x) |
| Onboarding time per hire | 4 to 8 hours | 45 minutes (about 85% reduction) |
| Security incidents | 2 in prior 12 months | 0 across 3 years post-build |
| Help desk FCR | 62% | 91% |
| Cost per user month | Baseline | Down 24% despite higher security coverage |
| Internal IT headcount | 1 lead | 1 lead, unchanged |
What the table doesn’t capture is the operational story behind it.
Fusion Computing’s monitoring caught the kind of pattern an internal IT
lead never has time to root-cause alone. Fusion Computing handled the
office relocation as a controlled IT project, not a cutover gamble.
The infrastructure was rebuilt before hiring velocity forced emergency
decisions, which meant the business never had to pause to re-architect
under pressure.
They didn’t replace my IT guy, they made
him better. Three years in, we’re six times bigger and IT isn’t even
on my list of risks anymore.CEO, Toronto design studio (anonymized)
The Canadian Centre for Cyber Security flags ransomware as the top
cyber threat to Canadian organizations, and the studio’s sector and
size profile (per Statistics Canada
SMB data) makes it exactly the kind of business that gets targeted
during fast growth, when controls usually lag headcount. The fact
that this studio went three years without an incident through a 6x
scale is the single number Fusion Computing values most about the
engagement.
What I’d tell another scaling-stage CEO
Build ahead of growth, not into a crisis. Having internal IT
isn’t the same as being covered. Co-managed only works when the
split between internal and external is explicit and written down.
Creative production environments need a different IT standard than
low-complexity offices.
The right outcome is never dependence on the MSP: a good
co-managed IT
relationship makes the internal IT person more valuable, not less.
That’s the test. If your current model is failing it, it’s time to
look at what a co-managed structure could actually do for your
business.
Frequently asked questions
Is co-managed IT only for larger businesses?
No. Co-managed IT usually starts to make sense when a business
has one internal IT person or a small team that’s stretched too thin.
It fits best when a company has enough support volume to justify
internal presence but not enough for deep in-house specialization.
Fusion Computing sees the strongest fit in the 30 to 150 user range,
where the internal lead is the relationship anchor and the MSP carries
the operational depth.
Would fully managed IT have made more sense for this studio?
No. The studio already had internal IT value worth preserving.
The better answer was to support that role, not remove it. A fully
managed model makes sense when there’s no internal IT team or when
leadership wants to outsource ownership entirely. This studio needed
partnership, not replacement. That distinction is the single most
common reason a co-managed engagement either succeeds or fails: it
has to match what the business already has.
How long does a buildout like this usually take?
It depends on procurement, site readiness, and how much
standardization is required. Fusion Computing typically scopes an
initial design and implementation phase of 3 to 6 months for a
30 to 100 user environment, followed by a stabilization period where
documentation, patching baselines, escalation paths, and support
rhythms get cleaned up and formalized. The studio engagement ran
4 months for the build and roughly 6 more for full stabilization.
What should a business expect from a co-managed provider?
Clarity. The provider should explain exactly who owns what, how
escalations work, what happens after hours, and how monitoring,
patching, security, and backup oversight are handled. If those
answers are vague, the engagement will get messy fast. Ask for the
written split of responsibilities before signing. A real co-managed
agreement looks like the table earlier in this case study, not a
generic SLA.
How did Fusion Computing handle cybersecurity while the user base grew six times?
Security was built into the operating model from day one rather
than treated as a future phase. EDR ran on every endpoint, Microsoft
Defender for Office 365 covered email, MDR monitoring ran against a
CIS Controls v8.1 baseline, and MFA was enforced on every account
that touched client data. Fusion Computing’s project records show
zero material security incidents at the studio across 3 years post-build,
against 2 in the 12 months before engagement.
What lessons from this engagement apply to a 30 to 100 user SMB planning similar growth?
Three lessons travel well. First, identity and device baseline
(SSO, MDM, conditional access) is the foundation everything else sits
on. Skip it and every later layer leaks. Second, templated onboarding
pays back the day a hiring plan accelerates: the studio’s 85%
reduction in setup time was the difference between IT being a hiring
constraint and IT being invisible. Third, the co-managed split has
to be written down. Verbal agreements drift into chaos.
Fusion Computing helps growing businesses scale infrastructure
without IT becoming the bottleneck across Toronto and the GTA,
Hamilton, and
Metro Vancouver. If
your business has one internal IT person carrying too much, a
co-managed model can add
operational depth, stronger security, and after-hours coverage
without replacing what already works. See related sibling case
studies for an internal lead supported by co-managed IT
or a multi-site cannabis retailer.
