What Is a Managed Technology Service?
Your email stops working at twenty to nine on a Monday. Your office manager calls the person who set up the network two years ago, gets voicemail, and sends a text. By noon you have lost most of a morning, three customers have emailed your personal address, and the fix, when it comes, takes fifteen minutes. The invoice is for two hours plus a trip charge.
Nothing in that story is unusual, and nobody in it did anything wrong. It is simply how break-fix support works. Something breaks, you find someone, they fix it, you pay for the time. A managed technology service is built on a different idea. Somebody is responsible for the condition of your systems on an ongoing basis, so that most of those Monday mornings never happen.
This article explains what that really means in practice, what you should expect in writing, and how to tell a careful provider from a hopeful one. It is aimed at an owner who is weighing the two models and wants to decide with clear eyes.
The plain definition
A managed technology service is an ongoing arrangement in which a provider takes responsibility for defined parts of your technology for a recurring fee. The word "defined" carries most of the weight. A good agreement says exactly which systems, which tasks, and which outcomes are covered.
Compare that with break-fix, where the relationship is transactional. You pay when there is a problem. The provider earns money when things break, which means their incentives and yours point in slightly different directions. That is not an accusation, only a structural fact. Under a managed arrangement the provider is paid whether or not something breaks, so prevention becomes the cheaper path for them.
Neither model is wrong for every business. A five-person office with a handful of laptops and a cloud email account may be well served by a trusted person on call. Once your operations depend on several connected systems, and an outage means lost sales or missed appointments, the math begins to change.
What is usually inside the scope
Scope is the first thing to read and the last thing most people read. Ask for a list, in plain language, of what is included. Typical categories look like this.
- Monitoring. Software watches servers, network devices, backups, and key services, and raises an alert when something looks wrong. Monitoring only helps if someone is assigned to respond to the alerts.
- Patching and updates. Operating systems and applications receive security updates on a schedule, with testing where it matters.
- Backups and recovery. Data is copied somewhere safe, and, importantly, restoration is tested. A backup that has never been restored is a hope, not a control.
- Identity and access. Accounts are created for new staff, removed for departures, and protected with multi-factor authentication.
- Email, domain, and DNS administration. Someone keeps the domain registered, the records correct, and the email authentication settings (SPF, DKIM, and DMARC, which together help receiving servers trust your mail) in good order.
- Help desk. Staff can ask for help and get a response.
- Vendor management. The provider talks to your internet service, software vendors, and hardware makers on your behalf.
- Reporting and planning. You get a periodic summary of what happened and what should be done next.
Equally important is what is not included. Projects such as migrating to a new system, rolling out a new office, or building a custom integration are often billed separately. That is reasonable, provided it is stated up front.
Response expectations, and why the word "response" is slippery
Most agreements describe service levels, which are promises about how fast the provider will act. The part that trips people up is the difference between response time and resolution time.
A response time is how long until a person acknowledges and starts working on your request. A resolution time is how long until the problem is fixed. A provider may promise a 30 minute response and say nothing about resolution, because resolution depends on the cause, and some causes lie with a third party.
A fair agreement usually sorts requests by severity.
| Severity | Example | What a sensible agreement states |
|---|---|---|
| Critical | Whole office cannot work, or a security incident is suspected | Fast response, ongoing contact until restored, available outside business hours |
| High | One important system is down for several people | Same business day response, with a target for restoration |
| Normal | One person cannot print or access a folder | Response within business hours, resolved in the normal course |
| Request | New account, new laptop, a question | Scheduled, with a typical turnaround |
The hours matter as much as the numbers. A promise of a fast response during business hours, Monday to Friday, is a very different product from one that covers evenings and weekends. If your clinic or shop runs on Saturdays, you need to hear Saturday in the agreement.
How tickets work, and why you want them
A ticket is a written record of a request. It has an owner, a status, a history, and a closing note. It sounds bureaucratic until the day you need it.
Tickets give you three things. First, accountability, since a request cannot silently disappear. Second, a memory. When the same printer fails for the fifth time, the history shows it, and the right answer is a new printer, not a sixth fix. Third, evidence for the periodic review, because you can see how many requests came in, which were repeats, and how fast they were handled.
Good providers make it easy to open a ticket by email, phone, or a portal, and they confirm receipt. Be wary of an arrangement that runs entirely on text messages to one person. It feels friendly, and it leaves no record.
Shared responsibility, the part nobody draws on a whiteboard
Here is where misunderstandings hide. Even with a provider, some responsibilities stay with you, and some belong to third parties.
Your provider may manage your accounts, and you still decide who should have access. They may set up backups, and you still decide how long records must be kept. They may protect your network, and an employee can still approve a fake payment request in an email. Technology cannot fully cover judgment.
Cloud vendors have their own version of this. The company that hosts your accounting software is responsible for keeping that service running and for protecting the platform. You are responsible for who logs in, what they can do, and what you do with the data. This split is commonly called a shared responsibility model, and every serious cloud vendor publishes one.
A useful exercise is to draw three columns on a page, one for what the provider does, one for what you do, and one for what a vendor does. Fill it in together at the start. Any task that lands in no column is a gap, and gaps are where outages and arguments come from.
Pricing models, and what each one rewards
Providers commonly use a few pricing structures. I will describe them without quoting market prices, because prices vary with scope, region, and the age of your environment.
- Per user. You pay a recurring amount for each person. It is easy to understand and scales with headcount. It can hide the effort required for unusual setups.
- Per device. You pay for each computer, server, or network device. It suits equipment-heavy businesses and can penalize you for having many small devices.
- Flat fee. One recurring amount for a defined scope. Predictable, though the scope has to be written carefully so that "flat" does not turn into endless extras.
- Tiered. Packages with increasing coverage. Easy to compare, with the risk that the feature you need sits one tier above the one you can afford.
- Hourly or block of hours. You buy time. Flexible, and it brings back the incentive problem if prices rise when there is more to fix.
Whatever the model, ask what is billed separately. Software licenses, hardware, and vendor subscriptions are usually passed through at cost or with a markup, and you should know which. For reference, Ozcorp InfraOps Base starts at $500 per month, with vendor and license charges billed separately. That sentence is the kind of clarity to look for from any provider. A starting figure, plus a clear statement of what sits outside it.
A worked comparison, using made-up numbers
Let us run a hypothetical, labeled as illustrative. A 20-person company has three outages a year that each stop about half the staff for three hours. Suppose the loaded cost of an employee hour, meaning pay plus overhead, is $40.
Each outage costs 10 people times 3 hours times $40, which is $1,200 in lost work. Three outages is $3,600. Add repair invoices of perhaps $400 each, and break-fix costs this company about $4,800 per year in visible terms, before counting missed sales or customer frustration.
Now suppose a managed arrangement cuts the outages from three to one, because patching and monitoring catch the causes early. The remaining outage costs about $1,600 including repair. The saving is about $3,200 per year. If the managed fee is higher than that, the fee is not justified by outages alone, and you would have to value other benefits such as security, planning, and fewer interruptions for staff.
The numbers are invented. The method is the point. Count the hours your people lose, multiply by what those hours cost, and compare it honestly to the fee. If the arithmetic does not favor a managed service, do not buy one.
Red flags to watch for
Years of reading agreements have made me suspicious of a few patterns.
- No written scope. If the provider cannot list what is covered, you are buying a feeling.
- Promises of "unlimited" everything. Unlimited support with no definition usually means unlimited disappointment.
- Admin accounts only the provider holds. You should always be able to get into your own systems. Ask who owns the domain registration, the cloud accounts, and the administrator passwords.
- No documentation handed to you. If you cannot leave without losing knowledge of how your own systems work, you are locked in.
- A long auto-renewing contract with no exit. Look at the termination terms and what help you receive on the way out.
- No reporting. If you never see what was done, you cannot judge whether it was worth it.
- Backups described but never tested. Ask when they last restored something, and ask to see the result.
- Security described only in adjectives. Look for specifics such as multi-factor authentication, patch cadence, and how incidents are communicated.
None of these proves bad faith. They are prompts for a direct question.
Questions to ask in the first meeting
- What exactly is included, and what is billed separately?
- What are your response targets, during which hours, and how are severities defined?
- How do requests reach you, and will I get a ticket every time?
- Who holds administrator access, and who owns the accounts and domains?
- How often do you test restores, and can I see the last test?
- How do you handle a security incident, and when would you tell me?
- What do I receive if I leave, and in what format?
- How many clients like us do you support per technician?
The last question tells you whether the person who answers your Monday call has the time to answer it.
A note on Puerto Rico and continuity
Hurricane season is part of the planning here, not an afterthought. Ask any provider how they would restore your systems if your office lost power and internet for a week. The good answer involves cloud-hosted services, backups stored off the island, documented recovery steps, and a plan for how your staff work from home or a neighbor's generator-powered connection. A thin answer is a useful result in itself.
Bilingual staff and customers deserve the same attention. Ask whether help desk requests can be handled in Spanish and English, and whether documentation is readable by the people who will actually use it.
What to do this week
You can learn a lot without calling anyone.
- List every system the business needs to operate, and next to each write who currently looks after it.
- Write down the last five times something broke, how long it took to fix, and what it cost.
- Find out who owns your domain registration and administrator logins. Write their names down.
- Ask whoever handles your backups when they last restored a file.
If the answers surprise you, you have found the real value of a managed service, whether you buy one or not. Knowing who is responsible for what is most of the job.
An invitation
If it would help to talk through your list, I am glad to do that. At Ozcorp Scientific we run managed services for small teams through Ozcorp InfraOps, and I will tell you honestly if what you have is better served by a good freelancer or a one-time project. Write to me with the five things that broke last, and we will start there.
Discuss your infrastructure requirements
Review the current service offer or schedule a conversation about the scope you need.
Stay in touch
Follow Ozcorp
Leave your details for occasional notes about new articles and Ozcorp updates.