ITIL v5's 8-Step Lifecycle, Explained With One Real M365 Automation
The ITIL v5 lifecycle made simple: I follow one Entra ID onboarding automation through all eight activities, from Discover to Support.
In my last post I summarised everything that changed from ITIL 4 to ITIL® v5. This week I'm going deep on just one part: the new 8-step product and service lifecycle. It's the backbone of v5, and one of the biggest topics on the Bridge exam.
My problem with it was simple. I could recite the eight names, but I didn't really get them. So instead of memorising definitions, I took one real thing I've built and followed it through all eight activities. That's when it clicked.
The example: an onboarding automation
Onboarding new starters used to be a manual job: create the Entra ID account, assign licences, add groups, send the welcome email. Automating it with PowerShell and Microsoft Graph is a very typical M365 project, which makes it a good test case.
The first thing v5 asks you to do is separate two things:
- The product: the PowerShell and Graph automation itself. It's something that gets designed, built, tested and maintained.
- The service: "automated employee onboarding" for HR. HR doesn't care about scripts. They care that a new starter has a working account on day one.
ITIL 4 mostly looked at the service. v5 manages both, and the lifecycle is the life story of the product and the service it makes possible.
Following it through all eight activities
1. Discover: should we build this, and why? HR tells us onboarding takes days and new starters sit around without access. Before writing any code, we check that automating it fits the IT roadmap and is worth the effort. Discover keeps what we build aligned with real needs and strategy.
2. Design: what exactly will it do? We sketch the flow: HR submits a form, the automation creates the account, assigns licences and groups, then sends a welcome email. Design covers the experience too: what does HR see, and what happens when a field is wrong?
3. Acquire: what do we need? An app registration with the right Graph permissions, the licences, somewhere to run the automation. Each of these is a resource the service depends on, so they get recorded, for example in the CMDB.
4. Build: does it work? We write the script, test it against a test tenant or test users, and ideally run the tests automatically every time the code changes. That's continuous integration.
5. Transition: how do we get it live safely? We release it to production, switch off the manual process and make sure the service desk knows what's changed. A bad transition means a broken onboarding on someone's first day.
6. Operate: is it running healthily? We monitor the automation's runs, alert on failures and keep the app's secret or certificate from expiring. This is about looking after the thing.
7. Deliver: are people getting the value? HR submits new starters, and the accounts are ready on time, as agreed. This is about serving the people: requests, users and service levels.
8. Support: what happens when it breaks? A run fails and a new starter can't sign in. We fix it fast, then look for the cause so it doesn't happen again.
What made it click
Three groups helped me. This isn't an official ITIL model, just how I think about it. Discover, Design, Acquire and Build are mostly about making the product. Deliver and Support are mostly about people using the service. Transition and Operate are the bridge between the two. Once I saw that, the order made sense.
It's not a waterfall. In real life the onboarding work loops. A Support issue might show HR needs contractors onboarded too, which sends us back to Discover. ITIL calls each real path through the activities a value stream.
Engage and Improve disappeared, but not really. In ITIL 4 they were steps in the value chain. In v5 they happen in every activity. We talk to HR during Design, Deliver and Support, and we improve the automation all the time. So they're no longer separate steps.
The pairs I kept mixing up:
| Pair | How I tell them apart |
|---|---|
| Operate vs Deliver | Looking after the thing vs serving the people |
| Event vs incident | A run finished (a change of state) vs a run failed and someone can't sign in (an unplanned interruption) |
| Incident vs service request | Something's broken vs HR asking for a normal, agreed thing, like onboarding a starter |
| Continuous delivery vs deployment | Always ready to release, with a person deciding when vs released automatically once tests pass |
What I'm taking into the exam
- The order, with my mnemonic: Do Dogs Always Bark To Our Delivery Staff? (Discover, Design, Acquire, Build, Transition, Operate, Deliver, Support)
- One question per activity: why? what? with what? does it work? how do we go live? is it healthy? are people getting value? what if it breaks?
- My three groups (a study aid, not official): product side, bridge, service side.
My tip if you're studying too: pick one real ticket or project from your own week and say which activity it belongs to. A password reset is Support. A new Conditional Access policy runs through Design, Build and Transition. It's the quickest way I've found to make the lifecycle stick.
ITIL® is a registered trademark of PeopleCert group. This post contains my personal study notes and interpretation, not official ITIL guidance.