Chapter 1 — What managed IT actually costs in Houston (2026)
Pricing varies more than most buyers expect. Four pricing models dominate the Houston market, and the differences matter.
Per-user, per-month (most common)
- ✓Fully-managed, 24/7 help desk + security stack + patching + backup: $165–$225 / user / month
- ✓Mid-market, business-hours-only help desk, basic security: $110–$160 / user / month
- ✓Gutted service (stay-away territory): under $95 / user / month
The $165–$225 band is where real MSPs land when the service includes a modern security stack (EDR/MDR, SIEM or managed detection, email security, phishing training, immutable backup, written incident-response plan). Anything meaningfully cheaper usually means one of those layers was cut, or the help desk is offshore.
Per-device, per-month
- ✓Workstation: $75–$130 / device / month
- ✓Server: $220–$400 / server / month
- ✓Network gear (firewall, switch, AP): $15–$40 / device / month
Per-device pricing fits businesses with lots of shared devices (warehouses, manufacturing floors, clinics with check-in kiosks). It gets messy when you layer security tooling that prices per user, so most modern MSPs default to per-user and handle devices as an add-on.
Co-managed (you have internal IT)
Co-managed runs 40–60% of fully-managed pricing because you already own the “tier one” layer. A good co-managed package in Houston is typically $75–$135 / user / month, and it should include the security stack and the on-call escalation path — the things your internal person doesn’t have time to do well.
Project and hourly
- ✓Senior engineer (server / cloud / security): $195–$275 / hour
- ✓Project engineer (deployments, migrations): $145–$195 / hour
- ✓After-hours or emergency multiplier: 1.5×–2× base
Hourly work is for bounded projects — a Microsoft 365 migration, a firewall replacement, a compliance gap assessment. Don’t run your ongoing IT on hourly; it’s how businesses get $40k surprise bills after a bad month.
Chapter 2 — Seven questions every Houston MSP should answer cleanly
If an MSP fumbles any of these, pay attention. They’re not trick questions — they’re the answers you need to evaluate whether the team can actually run your IT.
1. Where is your help desk, and who answers the phone?
You want a US-based help desk with named engineers you can reach. Offshore tier-one desks can’t solve most real problems, and you’ll spend the call trying to escalate past a script. If the answer is vague, that’s the answer.
2. What’s your average ticket resolution time?
A legitimate MSP tracks this and can show you recent numbers. “Same day” is not a number. “Our P1 average is under 30 minutes, P2 under 4 hours, P3 same business day — here are last quarter’s actuals” is a number.
3. Can I see a sample monthly report?
Monthly reports should show: tickets opened/closed, top resolvers, patching compliance, backup health, security events, and a written executive summary. If they hand you a ticket dump with no narrative, their account management is weak.
4. What’s your security stack? Name the products.
You should hear specifics: EDR (CrowdStrike, SentinelOne, Defender for Endpoint), email security (Microsoft Defender for Office 365, OpenText, Abnormal), backup (Datto, Veeam, Acronis with immutable storage), phishing simulation (Keep Aware, KnowBe4). If the stack is “we use industry-leading tools,” that’s a brochure sentence.
5. What happens if you get breached?
An MSP sits inside every client’s network. Ask: what is your own SOC 2 or equivalent posture? Who has admin access to your tools? Do you carry cyber-liability insurance that covers client damages? What’s your own incident-response plan? This is the question most MSPs duck.
6. Can I talk to three references in my industry?
Not just “references available on request.” Three named clients in your vertical, ideally ones with similar compliance obligations. A 10-year MSP should be able to produce these without hesitation.
7. What’s your onboarding process?
The honest answer is a 30–90 day program: discovery, documentation, security baseline, monitoring rollout, user training. Anyone promising “we’ll be fully live next week” is either skipping steps or already has a team that works exactly like yours (rare).
Chapter 3 — How to read a managed IT quote
Most MSP quotes obscure the things that will cost you later. These are the line items to find and the ones to ask about.
- ✓Per-user vs. per-device rate — and what's included at that rate
- ✓Security stack line items: EDR, email security, phishing, SIEM / managed detection, DNS filter
- ✓Backup + disaster recovery — including which workloads (M365 mailboxes, SharePoint, servers, workstations) and where the offsite copy lives
- ✓Monitoring and patching — is it 24/7 or business-hours?
- ✓Help-desk hours and after-hours policy — what's in-scope at 11pm on a Saturday
- ✓On-site visits — how many are included, how are extra ones billed
- ✓Project work vs. managed-service work — where's the line?
- ✓Exclusions — read the 'not included' list carefully; this is where surprise bills come from
Chapter 4 — Co-managed vs. fully outsourced
The right model depends on your headcount, your current IT situation, and the kind of problems you’re trying to solve.
Fully outsourced (no internal IT)
Best for businesses under ~40 employees or businesses where internal IT is one overworked person who wears another hat (office manager, controller, owner’s spouse). You want an MSP that handles everything: help desk, security, strategy, vendor management, on-site.
Co-managed (you have 1–3 internal IT people)
Best for businesses in the 40–250 employee range. Your internal person knows your business cold — they’re irreplaceable for tier-one support, line-of-business app knowledge, and relationships. What they usually lack is time, a modern security stack, and 2am coverage. A co-managed MSP fills those gaps and gives your internal IT real peers to escalate to.
Fully in-house with project augmentation
Usually only makes sense above ~250 employees, and only when you can staff a real team (help desk + sysadmin + security lead + strategy). Even then, most in-house teams augment with outside help for compliance work and specialized projects.
Chapter 5 — Contract must-haves
What separates a professional MSP contract from a liability waiting to happen.
- ✓Written SLA with real response-time commitments by priority (not just 'best effort')
- ✓Data ownership clause: you own your data, your backups, and your documentation — no hostage situations at term end
- ✓Exit clause: what happens at term end, who holds admin credentials, how long the transition assistance lasts
- ✓Security-incident responsibilities: who does what, who pays for what, what triggers the incident-response plan
- ✓Included vs. out-of-scope — clearly enumerated, not just 'standard IT support'
- ✓Compliance scope: if the MSP supports your HIPAA, CMMC, WISP, SOC 2, or PCI obligations, the contract should name it
- ✓Proof of cyber-liability insurance with enough coverage to matter
- ✓Change control: how rate changes are communicated and how often they can happen
About term length
Most real MSPs sign 1-, 2-, or 3-year terms. That’s not a retention trap — it’s because a large portion of your monthly cost is vendor pass-through (EDR licensing, email-security seats, backup storage, M365 licensing, fiber connectivity), and those vendors discount based on term commitment. A longer MSP term = better pricing on the stack underneath.
What you shouldscrutinize is the auto-renewal language. A fair contract auto-renews with 30 days’ written notice required to cancel before term end — which is standard. Evergreen auto-renewals with a 90+ day notice requirement and no opt-out window are the predatory ones. Read it.
Chapter 6 — How to switch MSPs without breaking anything
Switching providers is the part buyers fear most. Done properly, it’s a controlled 30–60 day project with zero downtime. Done badly, you lose a domain, a backup chain, or admin access to your own tenant.
Phase 1 — Before notice
- ✓Get a copy of your current documentation from your outgoing MSP (network diagram, password vault, vendor list, license inventory)
- ✓Identify every admin credential that exists on your systems — M365, domain registrar, DNS host, firewall, backup, EDR console, any MSP-managed portal
- ✓Confirm where your data lives and who owns the licenses (you, the MSP, or a reseller on your behalf)
- ✓Inventory your backups and verify the most recent restore actually worked
Phase 2 — Parallel operation (30–60 days)
- ✓New MSP joins as a co-admin on every system, does discovery, documents what's actually there (often very different from what's on paper)
- ✓Agent installs and security-stack rollout run in parallel with the old provider still supporting day-to-day tickets
- ✓Backups are verified restored to new MSP's environment
- ✓Run a tabletop of the first real incident: who gets called, in what order, by whom
Phase 3 — Cutover
- ✓Admin transitions: old MSP admins removed, new MSP becomes primary
- ✓Vendor licenses transferred (M365, backup, security licenses should belong to you, not the MSP — if they don't, this is where you discover it)
- ✓Domain + DNS transfer if applicable
- ✓Old MSP hands over documentation package in writing
- ✓New MSP confirms in writing that transition is complete
Chapter 7 — Red flags that predict expensive problems
Things that look minor during the sales process but reliably turn into six-figure mistakes later.
- ✓Help desk is offshore or 'partner-supported' (translation: reseller in a trench coat)
- ✓Can't name their security stack or waves you off with 'industry-leading tools'
- ✓No real SLA — just 'we respond quickly'
- ✓Price is materially below market (<$95 / user / month fully-managed in Houston is a red flag)
- ✓They own the licenses, the tenant, or the domain on your behalf
- ✓No written incident-response plan — theirs or yours
- ✓The MSA is short, boilerplate, or 'whatever our standard is'
- ✓References are 'available on request' and never actually materialize
- ✓They resell hardware or a specific software brand and it's a huge part of their revenue (reseller incentives, not IT incentives)
- ✓No SOC 2, Cyber Essentials, or equivalent posture of their own
- ✓After-hours coverage is 'emergency only' with no defined hours or escalation path
- ✓They're uncomfortable answering the 'what happens if you get breached' question
What to do this week
- Pull your current MSP agreement. Find the SLA section, the exit clause, and the auto-renewal language. Highlight anything you can’t explain in plain English.
- Audit who has admin access to your Microsoft 365 tenant, your domain registrar, and your firewall. Anyone on that list who shouldn’t be there is a risk — start with old MSP accounts.
- If you’re not sure when your last backup was restore-tested, ask your provider to restore one file to a different location this week and screenshot it. Unverified backups are the #1 reason ransomware recoveries fail.
- If you’re evaluating MSPs, pick three and send all of them the seven questions in Chapter 2. The answers (and how quickly they come back) will tell you more than any sales meeting.
How Mako helps Houston businesses
We’ve been running Houston business IT since 2001. Our average client stays with us 10+ years, every engineer is a college graduate and carries a TWIC® credential, and we operate inside the Westland Bunker (a Tier III data center) for clients who need colocation alongside managed IT.
The seven questions in Chapter 2 are questions we’ve been answering honestly for 25 years. If you’re evaluating us alongside other Houston MSPs, we’re happy to answer them in writing — and we’ll name three clients in your industry you can call.
