
Guide
Managed IT Services: What They Include and How to Choose a Provider
Managed IT services are an arrangement in which an outside firm, a managed service provider (MSP), runs agreed parts of an organization's technology on an ongoing basis under a written agreement. The usual core is monitoring, patching, a help desk, backups, baseline security and dealings with other technology vendors. The arrangement works when the agreement states the scope, hours, response targets, exclusions, data ownership and exit terms. Before signing, ask how the provider protects its own accounts on your systems: in May 2022, seven agencies from five countries, CISA among them, warned of an increase in malicious activity aimed at MSPs and expected it to continue.
Updated November 2026
What managed IT services usually include
The usual work falls into six blocks. A block missing from the contract has no owner.
- Monitoring: alerts on servers, computers, network equipment and backup jobs, so that failures can be caught before users report them.
- Patching: NIST SP 800-40 Rev. 4 describes patch management as identifying, prioritizing, acquiring, installing and verifying updates. Ask for all five steps, for operating systems, applications and firmware.
- Help desk: one place to report faults and request routine changes, such as a new user, with tickets that leave a record.
- Backups: the NIST Cybersecurity Framework (CSF) 2.0, in subcategory PR.DS-11, expects backups that are created, protected, maintained and tested. Testing means restoring, not reading a success message.
- Security: multifactor authentication (MFA), endpoint protection, firewall and email filtering, and reviews of administrator accounts. The cybersecurity baseline guide puts these controls in order.
- Vendor management: internet providers, software publishers, warranty claims, license renewals and the replacement of equipment that no longer receives security updates.
In-house, outsourced or co-managed IT
Organizations run IT in one of three ways, and may change model as they grow.
- In-house: your own staff run IT. This gives direct control and knowledge of the business, but one or two people cannot cover every skill, vacation and sick day.
- Fully outsourced: a provider runs IT under contract. It suits organizations with no IT staff, provided someone inside still approves changes, purchases and access.
- Co-managed: internal staff keep some duties, often user support and business applications, and the provider takes others, such as monitoring, patching or after-hours coverage. The contract must split the duties line by line, or tasks fall between the two teams.
A practical rule: keep IT in-house when the daily workload is specific to your business and fills a full-time role; outsource it when nobody can own IT day to day; choose co-managed when a capable internal team needs more coverage or skills. In every model, CISA's guidance for small businesses reminds leaders that security culture cannot be delegated.
What the agreement should state: scope, hours and response
- Scope: the users, devices, sites, servers, cloud services and applications covered, listed in an inventory attached to the contract and kept current.
- Hours: when the help desk answers, and which events, such as a server down or a suspected intrusion, are handled on nights and weekends.
- Response targets: how fast a ticket is acknowledged and worked at each priority level, with a written definition of every level.
- Exclusions: what costs extra or is not done at all, such as projects, office moves, new hardware, cabling or recovery after a security incident, and how that work is quoted and approved before it starts.
- Reporting: a regular report on tickets, patch status, backup results and open risks, reviewed with a named person on each side.
Security, data ownership and exit terms
Joint advisory AA22-131A, published in May 2022 by CISA, the NSA, the FBI and the cybersecurity agencies of the United Kingdom, Australia, Canada and New Zealand, lists what customers should require of an MSP in the contract. Its points include: who owns each security duty, such as hardening, detection and incident response; how and when the provider reports an incident; MFA on every provider account that reaches your systems; no reuse of administrator credentials across customers. Separately, it advises every organization to keep its most important logs for at least six months.
State that your data, configurations, documentation, administrator credentials and the licenses bought in your name are yours. For backups, record the two values defined in NIST SP 800-34 Rev. 1: the recovery point objective, the point in time to which data must be recovered, and the recovery time objective, how long a system can stay in recovery before the business is harmed.
NIST CSF 2.0 (subcategory GV.SC-10) asks that cybersecurity supply chain risk management plans cover what happens after a service agreement ends. Exit terms should set a notice period, the handover of data and documentation, the sanitization of copies the provider keeps, following NIST SP 800-88 Rev. 2, and the disabling of the provider's accounts. AA22-131A warns that this last step can be overlooked when a contract ends.
Questions to ask before choosing a provider
- Which of our systems would you manage, and which would you not support?
- How do your staff sign in to our systems? Is every account protected by MFA and limited to what you manage?
- How do you decide which patches go first? CISA advises organizations to use its Known Exploited Vulnerabilities (KEV) Catalog as an input to that decision.
- How will you prove that our backups restore, and how often will you test a full restore?
- Who covers our systems when the person who knows them is away?
- Which tasks do you subcontract, and to whom?
- What happens, step by step, if we end the agreement?
A provider that cannot describe its own access controls or its restore tests has told you something important. Organizations that handle U.S. defense contract information have extra requirements to check; the CMMC guide explains them.
Onboarding checklist for the first weeks
- Agree on the inventory: users, devices, network equipment, cloud services, applications and licenses.
- Move administrator credentials into a password manager your organization controls, then change shared passwords.
- Create named accounts with MFA for the provider's staff instead of sharing existing logins.
- Install the monitoring and patching tools and confirm that every device reports in.
- Run a full backup and a test restore, and record the recovery point and recovery time actually achieved.
- Check the basics CISA lists for small businesses: MFA enforced by technical controls, above all on administrator accounts; no administrator rights on users' laptops; disk encryption on laptops.
- List every technology vendor and contract with renewal dates, and agree who calls each one.
- After the first month, compare what was found with what the contract covers.
Warning signs in a managed IT relationship
- Backups are reported as successful, but nobody can show the result of a recent restore test.
- Provider staff share one login, or reuse an administrator password across customers.
- Patch reports are missing, or list the same outdated devices month after month.
- Computers run an operating system past its end of support with no replacement plan; Microsoft ended support for Windows 10 Home and Pro on October 14, 2025.
- Every request turns into a billable project because the scope was never defined.
- The provider avoids questions about leaving, or treats your documentation as its own.
Field notes: running IT through outages, heat and storms
Where power cuts, storms and heat are routine, a few points deserve their own lines in the agreement.
- Power: who maintains the uninterruptible power supplies, tests their batteries and sets servers to shut down cleanly before the batteries run out.
- Remote reach: monitoring goes blind when a site's internet link drops. Ask whether a backup connection is part of the design.
- Storms: what is done before a forecast storm, such as an extra backup and a shutdown list, and the order in which systems come back.
- Off-site copies: at least one backup copy far enough away not to share the same storm or outage.
- Equipment rooms: heat and humidity shorten equipment life; add temperature alerts to monitoring and see the structured cabling guide for the rooms themselves.
Frequently Asked Questions
What is the difference between managed IT services and break-fix support?
With break-fix support, you call a firm when something fails and pay for that job; nobody watches your systems in between. A managed service makes the provider responsible for agreed systems all the time, including preventive work such as patching, backup checks and monitoring. Break-fix can suit a very small office if someone inside reliably checks updates and backups.
Is a managed IT provider the same as a managed security service provider (MSSP)?
Not always. An MSP runs day-to-day IT, usually including baseline security such as patching, MFA and endpoint protection. An MSSP specializes in security services, for example managed detection and response or the review of security logs. Some firms offer both. Ask which security functions the agreement includes, who watches alerts outside office hours and who may act on them.
Who should own the passwords, data and documentation?
Your organization. Provider staff should work through their own named accounts, while you keep an emergency administrator account in a password manager you control. Network diagrams, configurations, asset lists and backup copies should be yours and delivered on request, not only when the contract ends. A provider that holds the only copy of your credentials can make leaving slow and costly.
How is a response time different from a resolution time?
A response time measures how soon the provider acknowledges a ticket and starts work. A resolution time measures when the problem is fixed. An agreement may commit only to response targets, because a fix can depend on things outside the provider's control, such as a software vendor or a replacement part. Ask for both to be reported by priority level.
Sources
- CISA, NSA, FBI and allied agencies, AA22-131A: Protecting Against Cyber Threats to Managed Service Providers and their Customers (May 2022)
- CISA, Cyber Guidance for Small Businesses
- CISA, Known Exploited Vulnerabilities Catalog
- NIST CSWP 29, The NIST Cybersecurity Framework (CSF) 2.0 (February 2024)
- NIST SP 800-40 Rev. 4, Guide to Enterprise Patch Management Planning: Preventive Maintenance for Technology (April 2022)
- NIST SP 800-34 Rev. 1, Contingency Planning Guide for Federal Information Systems (source of the RPO and RTO definitions)
- NIST SP 800-88 Rev. 2, Guidelines for Media Sanitization (September 2025)
- Microsoft Lifecycle, Windows 10 Home and Pro (end of support October 14, 2025)
About this guide
GlobalTech Corp. wrote this guide. The company is based in Hormigueros, Puerto Rico, works in IT and cybersecurity, and has over 25 years of experience. It serves businesses throughout Puerto Rico and also clients in Florida, remotely and, for the projects that require it, on site. It is a SonicWall Certified MSSP for Puerto Rico and a Dell Certified Reseller. Its IT services page lists the IT part of that work, from servers and hardware to PBX phone systems and UPS, and the page on managed IT services in Puerto Rico explains how that work runs for businesses on the island.
More guides: Technology Guides.