Why spreadsheets fail your supply chain

Supply chain planning software and Excel models work on averages — average demand, average lead time, average yield. But your supply chain fails at the extremes: a Chinese New Year shutdown stretching a 6-week lead time to 14, a demand spike in one region pulling inventory from another, a single-source component disrupting three assembly plants simultaneously.

When planners can only model the average case, they compensate with safety buffers — safety stock, safety lead time, dual sourcing — added at every tier without visibility into how they interact. The result is a network that is simultaneously over-stocked at some nodes and stocked out at others, with no clear picture of which buffer is actually protecting you and which is just costing working capital.

The core insight: Supply chain risk is not linear. A 10-day supplier delay does not cause a 10-day stockout — it can cascade through reorder cycles and cause weeks of downstream shortage. Simulation is the only tool that captures this non-linearity reliably.

What a supply chain simulation actually models

We build a discrete-event or agent-based model of your network that captures every entity — suppliers, plants, DCs, retail nodes — and every flow — orders, shipments, inventory positions, cash. The model runs thousands of replications, each with different demand realisations and disruption events, and produces a distribution of outcomes: fill rate, stock days, transport cost, stockout frequency.

🏭

Multi-echelon inventory

Central DC, regional DCs and store-level stock modelled together — finds the right level at each tier without double-buffering.

🌐

Network design

Compare 2-DC vs 3-DC vs direct-ship configurations. Model transport cost, lead time and service level simultaneously.

📦

Supplier risk & disruption

Introduce probabilistic disruptions — yield loss, port delays, capacity constraints — and see the true cost of single vs dual sourcing.

📈

Demand variability

Fit real demand distributions per SKU cluster, model promotions, seasonality and new product launches with realistic variability.

🔄

Reorder policy optimisation

Test (s,Q), (R,S) and min-max policies across SKU segments — find the combination that minimises working capital at target fill rate.

🚢

Mode & lead time trade-off

Air vs sea vs rail — model the inventory cost of long lead times against premium freight costs to find the economic crossover.

How the project runs

1

Network mapping & data collection

We map all nodes, transport lanes, lead times and current inventory policies. Demand history per SKU is pulled from your ERP. Gaps are filled with structured estimates and sensitivity tested later.

2

Model build & validation

The simulation model is built in AnyLogic or Simio. We validate it against 12 months of historical actuals — stock positions, order frequencies, fill rates — before running any scenario.

3

Scenario design & experiments

Together we define the key decisions under study — network redesign options, safety stock levels, supplier strategies. The model runs hundreds of replications per scenario and produces statistically robust output.

4

Recommendations & handover

You receive a written report with ranked recommendations, confidence intervals and sensitivity analyses. The model is handed over for ongoing use — scenario runner included so your team can test future decisions independently.

What you get at the end

Ranked
network configurations with cost and service level for each option
Per-SKU
safety stock targets with statistical justification, not rule-of-thumb
Live model
handed over with scenario runner — your team can reuse it for future decisions

The deliverable is not a slide deck — it is an executable model plus a written decision brief. The model can be re-parameterised as your network evolves, extending the value far beyond the initial engagement.

Simulation vs planning tools — what's the difference?

CapabilitySupply chain simulationERP / planning software
Demand variability✓ Full stochastic distributions per SKU✗ Average demand only
Supplier disruption scenarios✓ Probabilistic disruption events modelled✗ Not natively supported
Multi-echelon interaction✓ All tiers modelled simultaneouslyApproximate, tier by tier
Network redesign comparison✓ Any topology, side by side✗ Requires manual re-configuration
Fill rate confidence intervals✓ Statistical output with percentiles✗ Single-point output only
Reorder policy optimisation✓ Policy type and parameters optimised togetherPartial — fixed policy structure

Tools & technology

AnyLogic is the primary platform for end-to-end supply chain simulation — its agent-based and discrete-event hybrid engine handles multi-echelon networks natively. For inventory-focused studies we also use custom Python Monte Carlo models where speed and transparency are priorities. Simio covers production-side integration where the supply chain model connects to a manufacturing simulation.

AnyLogic Python (Monte Carlo) Simio Supply Chain Library AnyLogic Cloud

We have been AnyLogic users since the platform's early industrial release and maintain access to the full Supply Chain Library. All models are version-controlled and documented so your team can maintain them without our ongoing involvement.


Inventory simulation — stock set by variability, not averages

Supply chain modelling answers where inventory should sit across your network. Inventory simulation answers a sharper question: exactly how much? The standard safety stock formula — Z × σd × √L — is elegant and widely used. It assumes demand variability is normally distributed, lead times are independent of demand, stockouts don't affect future demand, and replenishment is instantaneous once the reorder point is reached. None of these assumptions hold in real supply chains.

In practice, demand is lumpy — a few large orders arrive unpredictably between periods of quiet. Lead times cluster around the average most of the time and then occasionally spike — and the spikes correlate with demand peaks, when you most need the stock. Stockouts create backorders that inflate the next order, creating bullwhip effects that a formula cannot see.

The result: Formula-based inventory policies are either over-stocked (because analysts add buffers to compensate for distrust in the formula) or under-stocked when the formula's assumptions break down during demand spikes. Simulation replaces the assumptions with actual distributions fitted to your data.

What inventory simulation answers

Safety stock levels

How much buffer stock is needed to achieve a target service level — 95%, 98%, 99.5% — under your actual demand variability and lead time distribution, not the formula's assumed normal distribution.

Reorder point & quantity

When to trigger a replenishment order and in what quantity — balancing order frequency (transaction cost), holding cost, and the risk of running out between the order and its arrival.

Supplier lead time risk

Which suppliers' lead time variability contributes most to your stock risk — and what additional buffer is required per supplier to maintain your service level target under their specific delivery pattern.

Seasonal & promotional demand

How much to pre-build before a demand surge — seasonal ramp-up, promotional events, new product launches — so that the surge doesn't exhaust safety stock before replenishment can respond.

Monte Carlo vs discrete-event: which approach?

ApproachBest forTypical use
Monte Carlo simulationPure inventory policy questionsSafety stock & reorder point when inventory is independent of operational flow
Discrete-event (Simio / AnyLogic)Inventory + operations togetherWhen replenishment interacts with warehouse capacity, receiving dock throughput, or production scheduling
System dynamicsNetwork-level bullwhip & policyMulti-echelon networks, demand amplification analysis across supply chain tiers
Spreadsheet formulaSimple, stable environmentsSuitable when demand is genuinely normal and lead times are stable — rare in practice

What data do we need for an inventory study?

  • Demand history — 12–24 months of order or consumption data per SKU, with dates and quantities (ERP or WMS export)
  • Lead time records — purchase order history showing order date and goods receipt date per supplier and SKU
  • Current inventory policy — existing safety stock levels, reorder points, and order quantities
  • Service level targets — your target fill rate or order completion rate by product group
  • Cost parameters — holding cost rate and order transaction cost, if a cost optimisation is needed (not required for service level studies only)
  • Planned changes — new suppliers, new SKUs, upcoming demand events, or supply chain restructuring plans
FAQ

Common questions about
supply chain & inventory simulation

Supply chain simulation builds a time-driven, stochastic model of your network — every order, shipment and inventory movement plays out with realistic variability. Unlike ERP or S&OP tools, it does not assume average demand or fixed lead times. It shows you the distribution of outcomes: best case, worst case, and everything in between. This makes it the right tool when variability, disruption risk or a network redesign decision is the central question.
Classical inventory formulas assume normally distributed demand and deterministic or normally distributed lead times. In practice, demand is lumpy — mostly quiet with occasional large orders — lead times vary by supplier and season, and stockouts affect future demand through backorders. Simulation replaces these assumptions with actual distributions fitted to your data, and runs thousands of periods to find policies that meet your service level target across the full variability you actually experience, not just on average.
The minimum useful dataset is: demand history per SKU (at least 12 months, ideally 24), current inventory policies (reorder point, order quantity, safety stock targets), lead time data per supplier and transport lane, and a bill-of-materials or product structure if multi-level. We can work with exports from SAP, Oracle, or any ERP. Data gaps are filled with structured estimates and sensitivity-tested explicitly.
ERP safety stock calculations use the standard formula with normal distribution assumptions. If your demand is actually lumpy, seasonal, or driven by large intermittent orders — which describes most B2B and industrial supply chains — the ERP formula is systematically wrong. It either overstocks (most common) or understocks when variability is higher than assumed. Simulation fits the actual distribution from your order history and finds the policy that genuinely achieves your service level target. The result loads back into your ERP as updated reorder parameters.
Practically unlimited, but the most value comes from segmenting SKUs into clusters (by demand pattern, margin, lead time) and modelling representative SKUs per cluster rather than every individual item. For networks with 10,000+ SKUs this is essential for manageable model scope and fast run times.
Yes — disruption modelling is one of the primary use cases. We parameterise disruption events probabilistically: a supplier goes offline for N days with probability P per year, a port adds X days of lead time during peak season. The model then shows the downstream stockout risk and the inventory buffer required to absorb each disruption at a given service level.
Yes. If production capacity is a constraint — and for many manufacturers it is — we connect the supply chain model to a production simulation so that plant throughput, yield loss and changeover scheduling feed directly into the downstream inventory picture. This is most common in food & beverage, pharma, and automotive supply chains.
You receive: the executable AnyLogic or Simio model with documentation, a scenario runner interface so your team can test new questions without needing a simulation licence, a written decision brief with ranked recommendations and confidence intervals, and a data dictionary mapping model parameters to your source systems. We also offer a half-day knowledge transfer session for your planning team.
Yes — the model is handed over with full documentation and a rerun interface. When new SKUs are added, new suppliers are onboarded, or demand patterns shift, re-run the distribution fitting on the updated data and the model generates new policy recommendations. Many clients schedule an annual rerun as part of their inventory review cycle, and an immediate rerun when a major supplier change or new product launch is planned.
Free consultation

Let's talk about your network

Tell us your supply chain challenge — network redesign, safety stock, disruption risk or reorder policy. We'll tell you honestly whether simulation is the right tool and what the project would involve.

Response within 1 business day
Full NDA available as standard
Remote delivery worldwide
Transparent fixed-scope proposal

Germany — Dresden

Anton-Graff-Str. 24, D-01309
dresden@simulatefirst.com
+49 (0) 351 30906020

Poland — Wrocław

ul. Powstańców Śląskich 5, 53-332
polska@simulatefirst.com
+48 75 6406434

We respond within 1 business day · NDA available · No spam, ever

By submitting, your data will be used solely to respond to your enquiry. Privacy policy.