Lora Cecere - Outside-in planning, the orchestrator role, and don't AI stupid
In this episode, Ben and Wim speak with Lora Cecere, founder of Supply Chain Insights and one of the most well-known independent voices in supply chain research. Over the past decades, Lora has published hundreds of research reports and several books studying why supply chains underperform and what the best companies do differently.
Lora explains her concept of outside-in planning: how 80% of the data surrounding supply chains goes unused, what demand latency really means, and why bidirectional orchestration across demand and supply is fundamentally different from conventional S&OP.
We explore the evolving role of the planner as orchestrator: why the understanding of planning has declined, how organizational design and governance matter more than technology choices, and what the difference is between a planner who pushes data through a system and one who asks whether the plan is actually good.
On AI, Lora is direct: chasing agents and agentics without rebuilding the data layer is "AI stupid." She explains why semantic reconciliation, ontological structures, and unified data models are the unglamorous but essential foundation, and why most companies are nowhere near ready to exploit what AI can actually offer.
The conversation wraps with practical advice on where to start the outside-in journey, the buy vs. build question, and what Lora would tell her 30-year-old self to read and do.Key topics covered include:
- Outside-in planning: definition, the five core concepts, and how it differs from conventional supply chain thinking
- Demand latency and why it is often three weeks to four months in most supply chains
- The Supply Chain to Admire framework: measuring performance on a balance scorecard of growth, operating margin, inventory turns and return on capital employed
- Why smaller and mid-sized companies can benefit from outside-in planning — and why customer intimacy matters more than a data science team
- The evolving role of the planner: from number cruncher to orchestrator and supply chain engineer
- Don't AI stupid: why layering agents on top of existing systems without rebuilding the data foundation creates false progress
- The semantic layer and ontological structure: why ship-to and ship-from data can't talk to each other, and what it takes to fix that
- Buy vs. build: why building in-house means missing the evolution of planning technology and domain knowledge
- Practical first steps on the outside-in journey: the cross-functional review of known unknowns
TIMESTAMPS:
(0:03) – Welcome and intro(0:48) – How Lora got into supply chain: from chemical engineer to analyst(2:35) – What is outside-in planning?(4:24) – Bidirectional orchestration across source, make and deliver(6:35) – The main goal of outside-in: balance scorecard over functional cost(8:48) – Supply Chain to Admire: how Lora measures supply chain performance(10:39) – Why optimizing for functional metrics throws the system out of balance(12:55) – Demand sensing vs. outside-in planning: clarifying the definitions(16:01) – Bidirectional orchestration vs. conventional S&OP(21:21) – You don't need a data science team — you need customer intimacy(23:56) – Companies have more data than they think — and don't use it(25:31) – From planner to orchestrator: what Lora's supply chain engineer article was really about(29:32) – Why the definition of supply chain has become too narrow(33:32) – Start with organizational design, not the system(35:38) – Don't AI stupid: chasing agents without fixing the data layer(42:16) – The semantic layer: not a one-time task, it needs to evolve(45:11) – Ontology is not just data cleaning — it's definitions cleaning(50:06) – Buy vs. build: why building yourself will make you miss the evolution(52:11) – The supply chain knowledge gap and why it worries Lora(53:28) – What to read, follow and do: Lora's advice to her 30-year-old self(56:25) – Where Lora gets her inspiration
Related Articles
