Skip to content

Scenario planning and simulation

On a spreadsheet, a site closure looks simple until the contracts are signed. A scenario model shows the consequences while the decision is still open.

  • Free 30-minute call
  • No commitment
  • You own the code
Typical industries
Supply chain, Manufacturing
Problem type
Resource planning
You receive
A documented model, run on your data

In short: A supply chain scenario plan tests a decision in a model, such as closing a site or changing suppliers. The model runs it through many ways the future could develop. It shows the likely range of cost and on-time delivery while the decision is still open.

The problem

Next Tuesday, the board decides whether to close a distribution centre. You have one slide. It shows today’s sites and transport routes next to the plan without that centre. Both are priced on an average month. The plan without it comes out cheaper. Now ask what happens if customers in that centre’s region suddenly order more, or if a key supplier delivers late. The slide has no answer.

Why every change to sites and supply routes is a bet

Closing a site looks like a local decision on a slide. So does moving a production line, switching a supplier or redrawing a service territory. Each one changes three things across the whole operation: what you pay, how much stock you hold and how fast customers get their goods. You live with the result for years. By the time costs rise, the contracts are signed, the production line is gone and customers have already waited for late deliveries.

Why the single base case fails

The usual tool is a spreadsheet. You model your current sites and the proposed set-up, then pick whichever is cheaper in an average month. That average month is your base case, the starting point you compare against. It measures both options against past demand. It says nothing about the demand to come. It also hides the peak weeks, because the average evens them out.

The flaw of averages, in numbers

Sam Savage gave this problem a name in Harvard Business Review in 2002: the flaw of averages. Plans based on average conditions usually go wrong. Take the site nearest to the closed centre. It can ship 100 pallets a day. After the closure, it gets 80 on a quiet day and 120 on a busy one, half the time each.

The average is 100, so the spreadsheet says the site fits exactly. In practice, every busy day leaves 20 pallets behind. If those orders were promised for the next morning, they’re late, whatever the quiet day after looks like. The average month shows no problem, while on every second day 20 pallets go out late. A simulation that plays the days one by one sees it at once.

What changes

Today, you price two set-ups on an average month, and the board picks the cheaper one. With our model, we test the closure before you sign anything. We run it through a jump in demand, a supplier failure and demand moving to another region. Each of these cases is a scenario: one possible future. You see which customers get their goods later, which sites absorb the load and what each case costs. The board still makes the call, but it knows what each scenario costs.

A simulation model of your operation

Scenario planning and simulation let you ask “what if we do this?” and get an answer you can defend before you commit. The answer comes from a simulation model: a working copy of your sites, goods flows and operating rules. You change the copy and watch what happens, and your business never notices. Mistakes in the model are cheap, so you can try as often as it takes.

What you get

We build your operations simulation model on your sites, goods flows and rules. Then we check that it reproduces your figures from the last few months. Then we test every proposed change with your past data and with many possible futures. Say the closure saves money in the base case but costs more when demand moves to another region. You learn that while the decision is still open, and not only when you’re paying for long transports into a region where you no longer have a site.

  • A model written around your own scenario planning and simulation rules.
  • Results measured in the numbers you already track.
  • It slots into how you already work, and you can check every answer it gives.

How it works

We write your resources, the work they have to cover and your rules into one model. You get a plan that says which resource does what and when. It’s the plan that fits your goals best, such as the lowest cost or the shortest waiting time. Every rule is kept at the same time, and that’s the part a spreadsheet can’t do.

Three kinds of what-if scenario

First we build the model of your operation. Then the proposed change runs through the futures that matter. Demand-shock scenarios test it against a sudden jump in volume, demand moving to another region, a new large customer or a sharp drop. Supply-disruption scenarios add a supplier failure, a closed transport route, a port delay or a raw-material shortage. They show which sites and customers the disruption hits in the end.

Operational-change scenarios model the change itself: the closed facility, the new shift pattern, the merged distribution centre. We run all three side by side. Then we compare them on cost in francs and on how many orders arrive on time, say 97 out of 100. You also see how well the operation copes with each disruption. That way one good-looking average can’t hide what you gain and what you give up. You choose knowing when each option stops working.

Static and Monte Carlo simulation together

We run two kinds of simulation on the same model. Static simulation replays what actually happened, such as last year’s orders with the new set of sites. Monte Carlo simulation is named after the casino in Monaco, because it works with chance. The computer draws many random futures, each with its own demand and delivery times. The first shows how the change would have worked in your past. The second checks it against what could come.

How many simulated futures are enough?

People ask this often. The rule is simple: run until the result no longer changes. Each run is a random result, so the average over all runs wobbles a little too. That wobble shrinks with the square root of the number of runs. Four times as many runs halves it, and a hundred times as many cuts it to a tenth. We add runs until the ranking of your options is stable. We show you the spread, how far the results range, along with the average.

Stochastic optimisation across all your sites

Demand and delivery times vary by chance. Stochastic optimisation is optimisation that works with that chance. It treats demand and delivery times as a range of possible values, each with its probability. Then it plans stock and capacity for that whole range. A spreadsheet usually takes the average demand and adds a fixed safety margin. Thousands of scenarios across many sites make a big computation. We split it into smaller parts and calculate them one by one, so you have the result before the board meets.

Supply chain and manufacturing decisions

Every decision about your sites sets the cost and delivery times your supply chain can reach. Inventory planning and emergency plans then work within those limits. Manufacturing runs the same risk. If you merge production lines or change shift patterns, work and stock move between sites. Network design decides which sites to run and where. It produces the proposal. Supply chain scenario planning tells you whether to trust it before you spend.

Common questions.

Something else? Ask us directly.

A handful of named scenarios, and many runs inside each one. Think of it as two layers. The first layer is the named scenarios that people can discuss. For example: the base case, a jump in demand, a supplier that fails and demand that moves to another region. The second layer sits inside each scenario: we calculate it many times, each time with slightly different demand and delivery times drawn at random. We stop when the ranking of your options no longer changes. You decide with the first layer: for each named scenario, you see which option is cheapest and what it costs. The second layer makes sure those costs aren’t down to luck.

A forecast tries to name the most likely future. Scenario planning takes a decision and tests it against several futures, including unlikely ones that would be expensive. You need both. Your forecast feeds the base case. The scenarios show in which situations your decision stops working and what that would cost. So you can choose an option that also works when things don’t go as forecast.

Mostly data you already have. Orders by customer, product and site, ideally for a year or more so the seasons show. Site capacities and running costs. Your regular transport routes, with freight rates and transit times. Your suppliers’ delivery times, and how often they deliver late. Missing months or incomplete freight rates are normal. We find them when we check the model against your figures from past months.

Sometimes it can, and if that’s the case, we’ll tell you. Standard software is built for an average operation, though. Yours runs on its own rules: your contracts, your exceptions, the things you know by heart. A custom model writes those rules in, and that’s where the better plan comes from.

You don’t need either to start. We build the model from the data you’ve already got, even if it lives in three spreadsheets and one person’s head. You get a system you run yourself. We train your team until they no longer need us.

You see a first result with the prototype, before you commit to anything bigger. We start with a free 30-minute call about your operation and the model that would fit it. Then we build a working prototype and run it on your numbers.

Tell us what you’re planning.

In a free 30-minute call with a specialist, you find out whether optimisation can help.

  • Free, no commitment
  • You talk to a specialist directly
  • A clear go or no-go