That MDM Guy. thatmdmguy.com

Field notes on modern device management

Hi, I’m Dario.
I manage fleets — and I write down what actually works.

Welcome to That MDM Guy. This is where I turn day-to-day work in the console — enrolments that won’t stage, profiles that won’t install, scanners that won’t behave — into clear guides you can follow the next time you hit the same wall.

No theory for its own sake. If it’s here, it’s because I’ve built it, broken it, or fixed it in a real fleet.

Illustration of a laptop, phone, tablet and rugged scanner connected to a central device-management dashboard with security checkmarks
ONE CONSOLE · MANY FLEETS — PHONES, MACS, TABLETS & SCANNERS

Hub notifications with no Target Audience: tracing it back to a Group ID mismatch in Access

You open Hub Services to send a notification — in my case, an outage notice — and the Target Audience Type field has nothing in it. No users, no groups, no way to send it to anyone. It looks like a permissions problem. In a production tenant, it usually isn’t the one you think it is.

TL;DR

If Hub notifications show no Target Audience Type options — and Hub Services shows no admins — check how Access points at your UEM tenant: in the Access console, under Integrations → UEM, the Group ID has to match the root Organisation Group’s Group ID in the UEM console exactly. In my tenant it held the OG’s friendly name instead of the real Group ID — close enough to look right, wrong enough that Hub Services couldn’t resolve who you are or who you’re allowed to notify.

1. The symptom

The notification editor itself loads fine. You can write the title and the message. Then you get to targeting — and stop:

  • Target Audience Type is empty. The dropdown/list offers no options, so the notification can’t be addressed to anyone.
  • Hub Services doesn’t list the admins you expect. People who administer the tenant in UEM don’t appear where Hub Services expects them.
  • Nothing about the message itself is wrong. This fails before content, before priority, before scheduling — at identity and audience resolution.

That combination matters. An empty audience on its own could be a scoping quirk. Empty audience plus missing admins points underneath Hub Services, to how it authenticates and looks people up.

2. What I ruled out first

Before touching Access, I checked the obvious UEM explanations — because if one of these is true for you, fix that instead:

  • Organisation Group scoping. I was working as a Global Admin at the top-level OG. This wasn’t a child-OG admin who couldn’t see the parent, and it wasn’t a scoped role missing a permission.
  • “Just give them a bigger UEM role.” Hub Services has its own admin model on the Access side. A bigger UEM role doesn’t automatically make someone a Hub Services admin — Hub Services Super Admin is a separate role in Access, and conflating the two sends you in circles.
  • Browser / session weirdness. A fresh session didn’t change it. Missing data that survives a clean login is configuration, not cache.

The lesson in the ruling-out

Hub Services looks like a UEM feature because you launch it from the UEM world. It behaves like an Access feature underneath. When the symptom is “people and audiences don’t exist,” think Access first.

3. How Hub Services is actually wired

If that middle link is wrong, the ends can both look healthy in isolation: UEM works, Access works — but Access is pointing at an Organisation Group it can’t actually resolve, so Hub Services can’t pull audiences or admins through and the fields that depend on that lookup come back empty. That’s exactly the failure shape here.

4. The fix: the Group ID in Access has to match the Group ID in UEM

My first suspect was the OAuth 2.0 UEM client in Access (Settings → OAuth 2.0 Management → UEM) — worth a look, and it checked out fine. The actual culprit was one screen over, on the Integrations → UEM page: the Group ID field.

Here’s the shape of it, with a generic company name swapped in. In the UEM console, the root Organisation Group’s real Group ID is in UEM’s own format — say acme_root. In Access, the Group ID field held the OG’s friendly, display-style name instead — say Root Acme. Same organisation, two different strings — and Access matches on the string, not on what you meant. Pointed at a Group ID that doesn’t exist in UEM, Access can’t resolve the tenant’s users, groups, or admins — so Hub Services shows no audiences and no admins.

Workspace ONE UEM console organisation group switcher for a fictional tenant, Acme, showing Group ID acme_root and the sub-organisation groups North, South and East — acme_root is the value Access has to match
Start in the UEM console. Open the organisation group switcher at the top of the console: it shows the OG’s real Group ID — acme_root in this dummy tenant. This is the source of truth — copy it exactly, underscores, casing and all.
Workspace ONE Access console, Integrations, UEM configuration page for a fictional tenant, with dummy values throughout — example API URL and API keys, account Alex Example — and the Group ID field, Root Acme, at the bottom of the configuration form
Then check Access → Integrations → UEM. The Group ID field at the bottom of the UEM configuration form is the one to compare — in my tenant it held the friendly name, not the Group ID from the screenshot above. Every value on this page is dummy data: real API URLs, keys, and certificate passwords are tenant-specific secrets and never belong in a post, a ticket, or a chat.
  1. In the UEM console, read the real Group ID.Open the organisation group switcher and note the root OG’s Group ID exactly as written — in the generic example above, acme_root, not the OG’s display name.
  2. Go to the Access console — not UEM.Sign in to Workspace ONE Access with an admin account and open Integrations → UEM.
  3. Compare the Group ID field, character for character.If it holds a friendly name (like Root Acme), a different OG’s ID, or anything other than the exact string from step 1, that’s your mismatch.
  4. Replace it with the exact UEM Group ID and save.Copy the value directly, console to console — don’t retype it from memory, and don’t “tidy it up” into a readable name. Readable is what caused this.
  5. Give it a clean test.Sign out, sign back in, and open Hub Services fresh before you judge the result — you want the corrected value evaluated in a new session, not a cached one.

The screenshots above are re-created with dummy data for a fictional company, Acme — no real tenant values anywhere — but the field names, the page paths, and the shape of the mismatch are exactly as they were. That’s the transferable part, and the part I wish a search result had told me up front: when Hub Services can’t see anyone, compare the Group IDs before you touch anything else.

5. Check the other half: Hub Services admins

While you’re in Access, close the second loop the symptom exposed:

  • Confirm the people who should administer Hub Services are actually listed as Hub Services admins — don’t assume the UEM admin list carries over.
  • Confirm at least one person holds the Hub Services Super Admin role in Access. That role is separate from UEM admin roles.
  • If the admin list is empty even for a known Super Admin, treat that as the same root-cause trail — identity resolution — not a second, unrelated bug.

6. Verify it worked

  • Create a new Hub notification (don’t just re-open the broken draft).
  • Target Audience Type now offers options — you can pick an audience type and resolve users/groups.
  • Send a test notification to a small, known audience first — yourself or your team — before the real outage or announcement message.
  • Confirm the test lands in Hub on a managed device before you rely on it for a fleet-wide notice.

If the audience still comes back empty after the client checks out, you’ve still made progress: you’ve eliminated the most common structural cause, and what’s left is a tenant-specific issue for support — with far better evidence than “the dropdown is empty.”

7. If you’re opening a case

Give support the shape of the failure, not just a screenshot of an empty field:

  • The exact symptom: no Target Audience Type options when creating a Hub notification, and no admins listed in Hub Services.
  • What you ruled out: admin level and OG (e.g. Global Admin at the top-level OG).
  • The Group ID on the Access side (Integrations → UEM) versus the Group ID in the UEM console: what each one said, and the change you made to line them up.
  • Whether Hub Services Super Admin is assigned in Access, and to whom (names/roles — never secrets).

That’s the packet that gets you past the first round of “check your admin role” replies — because you already did, and you can say so precisely.

Related reference: the MDM Almanac entry “Hub Services Group ID match” below is the one-line version of this post — the setting, where it lives, and what breaks without it. Start there next time; come back here when you need the full trail.

What you’ll find here

Three kinds of writing, all from live environments — not lab demos.

01

Field guides

Step-by-step builds you can follow in the console: cleaning up unused profiles, smart groups and empty organisation groups, modernising Apple enrolment through ADE, and moving a fleet from a shared tenant to a preferred tenant — iOS first, then macOS, then Zebra.

02

Troubleshooting notes

The short, honest version of a problem: what you saw, why it happened, and what fixed it. Like why a directory user set to Advanced — Enroll on behalf of another user never appears as a Default Staging User in an ADE profile — and why Standard staging fixes it.

03

The MDM Almanac

A searchable reference index beside the blog — settings, modes, parameters and error meanings you need in the middle of a build, without re-reading a whole article to find one line. See the preview below.

Topics

The platforms and problems this blog covers. Use the search above to filter them.

Workspace ONE UEM

Console

Profiles, smart groups, organisation groups, assignments and the cleanup work that keeps a tenant readable years later.

Apple ADE & ABM

Apple

Automated Device Enrolment, staging modes, dedicated vs. shared iPads, loaner setups, and why the OG on the profile has to match the user.

macOS management

macOS

Mac enrolment, Platform SSO, app cleanup, and moving Macs to a preferred tenant without breaking the people using them.

Android & Zebra

Zebra

TC78 fleets, Launcher / kiosk mode, MX and OEMConfig, and reading vendor XML line by line — including when the quotes in it are broken.

Hub Services & Access

Identity

Hub notifications, target audiences, and the Group ID in Access (Integrations → UEM) that has to match the UEM console before half of it works at all. Read Post 001 →

Tunnel & connectivity

Network

Moving from legacy Unified Access Gateway tunnels to Omnissa Tunnel containers on Docker, and checking what breaks if you get it wrong.

Shared & loaner devices

Fleet ops

Corporate-dedicated vs. shared, single-user staging, and loaner programs where the device — not a person — is the user of record.

Intune & tenant migrations

Migration

Shared-to-preferred tenant moves, gated phase by phase, and where Intune fits alongside — or instead of — Workspace ONE.

MDM Almanac — preview

The reference index that will sit beside the blog at launch. These are the first entries — the ones I look up myself.

EntryWhat it coversStatus
ADE staging modes

Standard (Single User) vs. Advanced — Enroll on behalf of another user, and why only Standard appears in an ADE profile.

At launch
AccessMgr — Zebra MX

PermissionAccessAction, granting MANAGE_EXTERNAL_STORAGE to Intelligent Hub silently on locked-down Launcher devices.

At launch
Hub Services Group ID match

The Group ID in Access (Integrations → UEM) must match the root OG’s Group ID in the UEM console exactly — a friendly name in that field means missing Target Audience options. Full walkthrough in Post 001.

In Post 001
Organisation Groups & Smart Groups

How OG placement decides what a user, profile and device can see — and why an ADE profile built at the wrong OG finds no one.

At launch
Platform SSO — macOS

What to decide before you turn it on: minimum macOS, shared Macs, and wipe vs. re-enrol for the existing fleet.

At launch
Legacy UAG → Tunnel containers

The migration path from legacy UAG tunnel to Omnissa Tunnel Containers on Docker, and the checks to run before cutover.

At launch

About the author

I’m Dario Ludovino Fernandez — the MDM guy behind this site. My day-to-day is enterprise device management: Workspace ONE fleets, Apple enrolment, macOS, and rugged Android devices in the field.

Recent work I’ll be writing up here includes driving a Workspace ONE cleanup and Apple ADE modernisation, a shared-to-preferred tenant migration running iOS → macOS → Zebra scanners, and the legacy tunnel move to containerised Tunnel — all under real security review, with real users waiting on the other side.

This blog is my notebook, made public. If a guide saves you an afternoon in the console, it did its job.

  • Sitethatmdmguy.com
  • FocusWorkspace ONE · Intune
  • FleetsiOS · macOS · Zebra Android
  • Signature topicsADE staging · MX XML · Hub
  • FormatGuides + Almanac reference

Start with the first post.

Post 001 is live above: Hub notifications with no Target Audience, traced back to a Group ID mismatch in Access — with console screenshots re-created on dummy data to prove it. It’s the template for what follows — symptom, what I ruled out, the wiring underneath, and the fix in order.

Read the Hub notifications post THATMDMGUY.COM — WORKSPACE ONE · INTUNE · APPLE · ZEBRA