Who Owns Scope? And What Happens When AI Makes It Worse

Most agencies know they have a scope problem. Projects come in over budget. Rework rates hover between 20% and 40%. Death marches are a regular occurrence, not an exception. And when you ask what went wrong, the answers sound the same: the client changed scope, the brief was unclear, we underestimated.

But the scope problem is not a discipline problem. It is a structural one. Agencies scope at the point of maximum ignorance — the start of the project — using the people least likely to be accurate: optimistic senior sellers operating under deal pressure. Then they hand that scope to delivery teams through briefings that reliably fail to transfer understanding. The result is a predictable sequence: a Plateau of Optimism, a Chasm of Despair, and a Death March. Every agency has lived this. Very few have names for why.

AI makes the scope problem worse, not better. AI accelerates production — it generates content, code, and deliverables faster than ever. But it doesn’t improve understanding. If you run AI into poorly-scoped work, you get more output, faster, in the wrong direction. The gap between how fast you can produce and how well you understand what you’re producing gets larger, not smaller.

This session examines who should own scope, when, and why the conventional answers are wrong. We’ll explore:

• The five broken assumptions that drive scope failure — and the corrected assumptions that start fixing it

• Why you need two scopes: one for selling and one for delivery — and why conflating them is the original sin

• The Ignorance Gap and why every project follows the same trajectory from false confidence to crisis

• Why senior people are systematically the worst estimators — and what the research actually shows

• How AI is an accelerant poured on a structural fire — and why better scope is the only firebreak

Presented by Jack Skeels, CEO of Better Company. Over fifteen years and more than two hundred agency engagements, Jack has helped leadership teams crack the structural problems they’d stopped believing were fixable — the chronic rework, the coordination drag, the culture issues that survive every initiative. The results are consistent: 20–50% faster delivery, 40% less rework, dramatically happier teams and clients, and a 94+ Net Promoter Score from the leaders he’s worked with. His approach is conversational, not prescriptive — he works alongside agency principals to make the structural changes that produce those numbers, drawing on a body of applied research that includes the award-winning book Unmanaged and ongoing work on how AI is reshaping knowledge-work organizations. This is a 45-minute session with Q&A afterwards.