Model the system before optimising it
First find the loops and delays that produce the problem. Then decide what to change, and where.
Supply chain resilience and network design
Your supply network was designed for a forecast that will not happen.
I help you see where it breaks, what flexibility is worth buying, and which decisions to make now versus later. The analysis comes as an interactive tool your team keeps, not a slide deck.
For COOs, supply chain leaders and CFOs in mining, food and beverage, maritime and ports, and healthcare logistics.
Fig. 01 — Bullwhip loop, simplified
Two loops, one dysfunction: rising orders read as rising demand (R1), while the loop meant to rein them in (B1) only acts after a delivery delay — by which point the overshoot is already on order. This is the structure, not any one person’s forecast error.
The problems I get called in for
Fig. 02 — Problem, cause, evidenceWhat you are hearingWhat is usually going onSee it working
“We fix the same supply problem every year.”
Recurring problems come from feedback and delays, not bad luck.
See itSupply Pulse · Lag Lab
“Planning keeps calling it forecast error.”
Overshoot after a demand step is the structure of planning, not the forecast.
See itLag Lab
“We don’t know which disruption would really hurt us.”
Exposure is time-to-survive versus time-to-recover, not a risk-register score.
“We’re committing capital on one forecast.”
Flexibility has a price you can calculate. Stage, defer or lease before you commit.
“Commercial, supply and network decisions move at different speeds and collide.”
Clockspeed mismatch is a design problem, not a coordination problem.
See itClockspeed Balancing
How I approach it
Fig. 03 — Principles
01
02
Test every decision against a fair baseline
A response only counts if it beats what a competent team would have done anyway. Otherwise the gain is an illusion.
03
Value flexibility explicitly
Buffers, second sources and staged capacity cost money every year. They should carry a value in the same calculation.
04
Build a tool your team keeps
The model stays with your analysts. They can rerun it when the next question arrives, without calling me back.
Ways to work together
Fig. 04 — Start small, go deeper
Offer 01 / Start here
Disruption Diagnostic
For: COOs and supply chain leaders after a disruption, facing a problem that keeps coming back, or answering the board’s “what if X closes?”
Format: A 2-day executive workshop, with about two weeks of preparation and write-up. Fixed fee.
You leave with: A causal map the team agrees on, a ranked list of exposures, the three leverage points worth acting on, and a short decision memo.
Offer 02 / Go deeper
Network Flexibility Study
For: CFOs and COOs about to commit capital: a new site, a capacity expansion, a second source or contract manufacturer, or a network redesign.
Format: 8–12 weeks. Fixed fee per phase.
You leave with: The option set with a value on each option, a staged plan with clear decision triggers, a board-ready investment memo, and the model itself.
Offer 03 / Keep the capability
Decision Tool Build and Handover
For: Heads of supply chain strategy and planning who want the capability in-house, not another report.
Format: 6–10 weeks to build, then three months of fixed support.
You leave with: A working, validated tool with its code, a validation report, documentation and a team that can use it. Optional annual model review.
Most engagements start with the Diagnostic. I work with organisations in mining, food and beverage, maritime and ports, and healthcare logistics.
See it working
Fig. 05 — Live tools
Global Disruption Atlas
If a strait or a major port closes, which of our sites run out first, and what does each response cost?
It shows which buffers, second sources or air bridges are worth funding, by comparing each site’s time-to-survive with its time-to-recover.
Supply Pulse
Why does a demand peak or a plant stoppage turn into empty shelves, then written-off stock?
It tells you whether overtime, faster quality release or a different planning policy actually pays, measured against a fair baseline.
Tierwise
Which site in our multi-tier chain runs out of capacity first, and when must we add more?
It sizes the capacity to add, site by site and year by year, and shows how much demand is served if you add nothing.
Real Options for Network Design
Should we build full scale now, build smaller and expand, wait, or lease?
It puts a value on each option against the best fixed design, and gives the demand levels that should trigger the next decision.
The examples run on illustrative data, mostly from food and dairy. The methods transfer to other sectors.
PhD, Technical University of Denmark · MSc, MIT · Council member, System Dynamics Society · Published in the Journal of Simulation, the NATO Science Series and Computers & Security · Past sector experience in mining, beverage distribution, healthcare and pharmaceutical manufacturing · Copenhagen
Next step
Book a 30-minute conversation about your network
Technical depth → Validation and test suites, methods, research, the full tool portfolio and the lab.