Working With a Managed Service Provider

These are not projects with an end date; they are obligations that persist whether or not you have someone available this week.

Why managed services grew

Demand for managed service providers has risen dramatically over the past decade, and the reason is fairly mundane: running IT infrastructure well requires a breadth of specialist skill that most organisations cannot justify employing in full.

Security, networking, cloud platforms, backup, monitoring, service desk — each needs current expertise, and each needs cover when the person who holds it is on leave.

The failure mode is a provider meeting every SLA while the service quietly deteriorates — because responsiveness was specified and improvement was not.

What an MSP actually takes on

The core of it is risk reduction around IT operations. In practice that means monitoring and alerting; patching and updates; backup and recovery; security operations; service desk and end-user support; and capacity and performance management.

The common thread is that all of it is continuous. These are not projects with an end date; they are obligations that persist whether or not you have someone available this week.

SaaS and managed services are not the same thing

The distinction is worth being precise about, because they are often conflated.

SaaS is software delivered as a service — the vendor runs the application, you use it. Your responsibility ends at configuration and your own data.

Managed services means someone operates infrastructure or applications on your behalf, including things you own. The scope is defined by contract rather than by product.

You can consume plenty of SaaS and still need managed services, because somebody has to run the identity, network, integration and endpoints that sit underneath it.

The benefits, and the honest caveat

The benefits are predictable cost, access to skills you could not retain in-house, continuous cover, faster response, and infrastructure spend reduced by not sizing everything for a peak.

The caveat is that an MSP is only as good as the contract and the relationship. The failure mode is a provider meeting every SLA while the service quietly deteriorates — because responsiveness was specified and improvement was not.

Choosing one

Ask what happens at the boundaries: who owns a problem that spans your systems and theirs. Ask what proactive means in the contract, and what evidence you will get that it happened. Ask how you would leave, and what you would be holding on the day you did.

The answers to those three tell you more than any capability matrix.