Skip to content
← All guides

Planning a deployment, step by step

Working out what you need, writing it down before you install anything, enrolling with consent, verifying it actually works, and reviewing it as circumstances change.

Practical9 min read

Most failed setups fail in the same way: someone starts by installing something. They pick a tool, enroll a machine, and work out what they actually wanted somewhere around step four — usually after a policy has already annoyed someone enough to get it removed.

This is the order that avoids that. It applies whether you are configuring one Mac for yourself or ten for a ministry.

Step 1 — Work out what you are actually trying to do

Before touching any software, get concrete answers to these. Vague answers here produce configuration that satisfies nobody.

  • Which devices?List them. Include the ones you cannot manage — a work laptop, a phone on someone else's plan — because those are where a policy leaks.
  • Who uses each one? An adult who asked for this, a teenager, and a shared machine in a church office are three different situations with three different answers.
  • What standard, specifically?“Block bad stuff” is not implementable. “Block adult content categories, allow social media, no internet after 11pm” is.
  • Is reporting wanted, and by whom? If yes, that person needs to agree to receive it before anything is configured.
  • Who owns the hardware, and who decides? Name this person explicitly. It prevents the situation where someone under a policy can quietly lift their own restrictions.

That last one is worth dwelling on. If the person being protected is also the person who can change the policy, you have a diary, not a safeguard. That may be fine — a lot of people genuinely just want the friction — but decide it deliberately rather than by accident.

Step 2 — Write it down before you install anything

Write a plain-language document, even if you are doing this for yourself. It should state:

  • Every restriction that will be applied
  • Every category of data that will be collected, and its retention
  • Everyone who will be able to see that data
  • Any third-party service involved, and what it costs
  • The process for requesting a change later
  • How to unenroll, stated up front rather than discovered later

Writing it down does two things. It forces you to notice decisions you were making implicitly, and it gives the person using the device something concrete to agree to. “I set up some filtering” is not consent. A page they read and agreed to is.

Step 3 — Enroll with consent

The person who uses the device should read the plan and agree before enrollment. Where the device is used by a minor, the parent authorizes it and the child is still told what is being applied, in terms appropriate to their age.

Mechanically, enrollment means:

  • Choosing a management service and setting up an account
  • Deciding whether you need supervision — see the supervision guide, since this determines whether the device must be erased first
  • Installing the enrollment profile and confirming it applied
  • Showing the device user where to view active profiles in System Settings, so nothing about the arrangement is mysterious

Step 4 — Verify it actually works

This is the step everyone skips, and it is the one that separates a deployment that works from one that merely installed. A profile reporting success only means the settings were accepted, not that the policy does what you intended.

On the actual device, confirm:

  • Blocked categories are in fact blocked — test a few
  • Sites the person genuinely needs for work or school still load. Finding this out during a deadline is how policies get removed.
  • Restrictions hold in a second browser, and in a second user account
  • Everything survives a restart
  • If reporting is enabled, a report actually arrives at the intended recipient
  • If you expect removal alerting, try removing a profile and check that the alert fires

Step 5 — Review it as things change

Circumstances move. A new job needs a category unblocked, a macOS release changes how a setting behaves, a fifteen-year-old turns eighteen. A configuration nobody revisits slowly stops matching real life, and the usual outcome is that someone removes the whole thing in frustration.

  • Set a recurring date to look at it — quarterly is plenty
  • Handle false positive reports quickly, or the policy loses credibility
  • Check compatibility before major macOS upgrades
  • Keep a note of what changed and when
  • Loosen policy deliberately as circumstances warrant

A realistic sense of effort

For one Mac, without supervision: an evening, most of it spent on steps one and two rather than on the software. Adding supervision means backing up, erasing, preparing the device, and restoring — a Saturday, comfortably.

For a group of devices, the per-device time drops sharply after the first, since the policy is written once and assigned to a group. The dominant cost is the conversation with each person, not the configuration.

If that sounds like more than you want to take on, the honest advice is to do step one and step two anyway. Even if you end up using a simple consumer tool, knowing exactly what you want and having written it down puts you ahead of most people who try this.