Here's a diagnostic you can run in your next steering meeting, and it takes ten seconds. Ask: "Who can kill this project?"
If the answer is a name, said quickly, you're probably fine. If the answer is a pause, a glance exchanged, or — the classic — three people answering at once, you've found the real risk on this initiative, and it isn't on the risk register. After five decades of delivery work between us, across startups and Fortune 500s, we'd put it plainly: most project failure is role failure. Not bad methodology, not bad software, not even bad luck — three jobs that were never clearly assigned, drifting until the drift becomes the crisis.
The three roles
The three jobs have many titles across frameworks. Strip the vocabulary and they are:
The Sponsor owns the WHY — and the money. The senior person whose goals this project serves. They fight for budget, remove obstacles the team can't, make the calls nobody else has authority to make — and they can kill it. A project without a real sponsor isn't underpowered; it's an orphan, and organizations starve orphans slowly.
The Owner owns the WHAT — and the benefit. The person whose part of the business changes, who will run whatever gets delivered, and who is on the hook for the result showing up after go-live. Not the person building the thing — the person who has to live with it. When benefits fail to materialize, it's almost always because this role existed only on a slide.
The Lead owns the HOW — and the plan. The project or program manager: sequencing, coordination, surfacing problems early, telling the truth about status. Crucially, the Lead owns delivery of the thing — not the business result, and not the decision to proceed. Those belong upstairs.
The absentee sponsor
Enthusiastic at kickoff, then vanishes into their calendar. The tell: the Lead starts escalating problems sideways — to peers, to steering decks nobody reads — because up leads nowhere. Decisions queue for weeks. The project doesn't die; it ages, which costs more. If you sponsor something, you owe it thirty minutes a week and your presence when it's in trouble. If you can't afford that, you can't afford the project.
The owner in name only
A department head "owns" the initiative, meaning: they attended the kickoff. The system launches, adoption stalls, benefits don't come — and everyone's surprised, though the one person accountable for the outcome never touched the work. The tell is early and visible: an Owner who delegates every design decision to the project team is telling you now the benefits case has no guardian.
The lead as scapegoat
The most common and least fair. Sponsor absent, Owner nominal — so the Lead, the only person visibly working the project, quietly absorbs both jobs. They start making scope calls that need sponsor authority and benefit promises that need owner conviction. It works until it doesn't, and when it doesn't, the failure wears the Lead's name — though the actual failure happened months earlier, when two other chairs went empty.
Three artifacts, three signatures
Three artifacts keep the three jobs honest, and the mapping is exact: the business case is the Sponsor's signature. The benefits plan is the Owner's. The delivery plan is the Lead's — and only that one. When one person signs all three, you don't have governance; you have hope with a Gantt chart.
The small-company translation
At 20 people you don't have three spare executives, and that's fine — three jobs, not three headcount. The Owner and Lead can merge when the person running the change also runs the affected area; it's workable. The merge that isn't workable: Sponsor and Lead. The person doing the work cannot also be the independent authority deciding whether the work should continue — that's how zombie projects survive.
The fix costs fifteen minutes at kickoff. Three questions, answered with names and said aloud to everyone: Who can kill this, and do they know they're expected to stay close? Who is measured on the benefits a year from now, and did they write the number? Who owns the plan — and is that all we're holding them accountable for?
Most stalled projects we're asked to rescue could have skipped the rescue with that one conversation held early — which is why it's the first exercise in our delivery programs at EZ Academy.