Scenarios

A question, written as levers

A scenario is an ordered list of levers, each with a value and the month it lands. That is the whole definition. Twelve of them ship ready made because they are the questions people actually ask, and every one of them is editable.

Build a twin Every lever, in detail

Every scenario is run against the same twin doing nothing, on the same seeds.

The interesting part of a decision is rarely the decision. It is the month it lands and what else is happening that month.

A price rise in month one and the same price rise in month six are different decisions, because in between them a competitor has finished matching your last move, two contracts have come up for renewal, and the sales team has either made its number or started discounting. The lever is the same. The world it lands in is not.

Anatomy

What a lever is

Three fields and sometimes a fourth. A key that says which lever, a value in whatever unit that lever uses, and a start month. Some levers carry options: which department to cut, how many months an integration takes, how likely a launch is to land.

Values are clamped to the range the lever allows, so a scenario cannot quietly ask for something the model has no equations for. A scenario with several levers applies them in the month each one is set to, and two levers can land in the same month.

Key: which lever, from the seven groups below
Value: a percentage, a count, an amount of money or a score, depending on the lever
Start: the month it lands, counted from the first month of the run
Options: the extra detail a complex lever needs, such as which department or how long an integration runs

Seven groups

What you can change

Group	Levers	What the engine does with them
Price	Change our price. Change price for new customers only. Allow sales to discount up to.	New and renewing customers see a price change now, contracted revenue sees it when its term ends, and a discount authority gets spent by a team that is behind quota.
People	Add or remove salespeople. Add or remove staff. Eliminate a department. Cut headcount by a percentage. Change pay.	Cost moves immediately, capacity and quality move with a lag, severance lands in the month of the cut, and morale carries the shock for months afterwards.
Supply	Supplier cost change. Switch to another supplier. Move manufacturing or delivery.	Cost of goods moves by the share that is actually the supplier's, and a switch or a move adds a disruption window where quality drops while the new arrangement settles.
Customers	Lose a named customer. Win a named account. Spend on keeping customers.	A named account is removed at that month with its revenue and its reference, a win adds revenue and the delivery load that comes with it, and retention spend buys satisfaction which buys retention, slowly.
Growth	Change marketing spend. Launch a product. Enter a new market. Acquire a competitor. Invest in the product or service.	Marketing buys pipeline with diminishing returns, a launch costs monthly until it lands and then lifts demand only if it hits, a market entry ramps from a standing start on brand, and an acquisition buys revenue and inherits a company.
Money	Raise capital.	Cash in, runway out, and an investor whose patience is now running on a clock.
Market	A competitor moves first. A competitor launches against you. Demand shock. A rule changes.	Things that happen to you rather than because of you. Your price did not change, but what your customers are comparing it against did.

A custom scenario can combine any of these. The ready made ones are simply combinations somebody has already written down.

Timing

Start months are where the second order effects live

The ready made scenarios do not all start in month one, and the choices are deliberate. Losing the largest customer starts in month three, so you can see the baseline first and the hole second. A competitor launch lands in month two and their price move follows in month three, because that is the order those things actually happen in.

An acquisition starts in month two so the purchase price and the integration drag are separable in the cash line. Moving manufacturing starts in month two so the disruption window sits inside the horizon rather than running off the end of it.

Change any of them. The most useful version of a scenario is usually the one where you have moved the start month to when the decision would really be taken.

Adaptive management

One switch that changes what the question means

Every scenario carries a flag for whether your own executives are allowed to react. It defaults to on.

With it on, the company answers back. Runway gets short and the finance lead cuts. Churn runs hot and the chief executive funds retention. The field is behind and starts discounting. Utilisation is high and operations hires. None of that is in your lever list, and in a lot of scenarios it is most of the answer.

With it off, you see the mechanical response on its own: what happens to this company if nobody at the top does anything about it. That is a useful and quite uncomfortable number.

The best use of the switch is to run the same scenario both ways. The gap between the two is a measurement of how much your answer depends on your own management reacting the way the model thinks they will.

What management does and when

The twelve

Each one has a page on what the model does with it and where it is weakest

These ship with levers already set and start months already chosen. The pages below describe the mechanism, not a promised result.

Raise price 20 percentMargin in month one, the customer answer over a year, the competitor answer somewhere in the middle.Cut price 10 percentVolume has to make up the margin, and competitors get a vote on whether it does.A competitor launches against usNothing you did changed. Everything your customers are choosing between did.Enter a new marketCost now, customers later, a standing start on brand, and the question of who is looking after the old market.Lose the largest customerConcentration is a number until the month it is an event.Acquire a competitorYou buy revenue and inherit a company. The second part is where the money goes.Eliminate a departmentThe cost line is easy. What they were quietly holding up shows up in month three.Hire ten salespeopleCash out for two quarters before the first of them clears quota.Our biggest supplier raises prices 15 percentMargin, then the decision about passing it on, then the customer answer to that decision.Move manufacturingUnit cost falls on a spreadsheet and quality wobbles in the real world.Launch this productA build window you pay for and a demand lift you hope for, modelled as a chance rather than a plan.Cut fifteen percent of the companyThe saving is certain. The second wave of leavers is the part that gets missed.

The baseline

Change nothing is a scenario too, and it is the important one

Every twin gets a scenario called change nothing, with an empty lever list. It is not a placeholder. It is the thing every other answer is measured against.

When you run a scenario, the same twin is run twice: once with your levers and once with none, on the same seeds, drawing the same parameters. The difference between the two is the decision. Without that, you are reading the noise in your own assumptions and calling it a result.

Why one baseline matters

Sweep the number you were arguing about

Twenty percent was a guess. A sweep runs the whole range and comes back with a curve and a best point.

Sweeps

Put the options side by side

Several scenarios, one baseline, one set of seeds, one table you can take into a meeting.

Comparing scenarios

Read what it rests on

The brief lists the assumptions carrying the answer, defaults first, because those are the ones worth replacing.

The brief

Pick the one you have been putting off

All twelve are already set up on the invented example company, so you can watch one run end to end before you upload anything of your own.

Build a twin See it on your numbers