Section 1
The artifact, part one: the zero-based contribution margin rebuild
Contribution margin is the plainest number in your business. It is what one unit of sale leaves behind after the costs that move with that sale are paid. Revenue per unit, minus variable cost per unit. Everything above your fixed costs has to come out of the pile of contribution margins you stack up in a month. When a policy shock hits, that per-unit number is what actually moved, and last year's version of it is now a lie. Rebuild it in five steps. Do this on paper or a single spreadsheet tab. Do not import last year's model and edit cells, because the anchoring will drag you back to the old shape. Start empty. Step 1. Define the unit. Pick the thing you sell one of: a covered table, a booked job, a service hour, a delivered order. Everything else in the worksheet is "per one of these." If you cannot name the unit, you cannot rebuild the margin, and that is its own finding. Step 2. Write the revenue per unit at today's real price. Not the list price. The price you actually collect after the discounts and comps you habitually give. If your average table pays 42 and not the menu's 55, write 42. Step 3. List every variable cost per unit under the new rules, from scratch. This is the step the shock breaks, so slow down here. For each cost that moves with the unit, write its value under the new policy, not the old one. Labour per unit gets recalculated on the new employment cost, not last year's. Materials get the current landed price including any new duty. Card fees, delivery, packaging, the lot. The discipline is to ask of every line: has a rule changed how this behaves, and have I entered the post-shock number? A cost that used to scale with wages but now applies per head has to be entered per head. Step 4. Compute the new contribution margin, per unit and as a percent. Revenue per unit minus the rebuilt variable cost per unit. If the number is smaller than you assumed, good, you found it before your bank did. Step 5. Recompute breakeven on the new margin. Fixed costs for the period, divided by the new contribution margin per unit, gives the units you must sell to cover the base. Compare that to what you actually sell. The gap between the breakeven you thought you had and the one you actually have is the size of the problem, stated in units you understand. That is the whole rebuild. It is deliberately boring. The point is not sophistication. The point is that you are now looking at the business that exists after the shock, not the one that existed before it, and every decision from here is made on real coordinates.
Section 2
The artifact, part two: sequence the levers by speed-to-cash and reversibility
The rebuild tells you the hole. It does not tell you which order to pull levers to fill it, and order is where operators die. A lever that fixes the margin but takes four months to reach the bank is useless if you run out of cash in ten weeks. So score every lever you are considering on two axes, and pull them in the order the grid gives you. Axis one: speed-to-cash. How many days until this lever actually changes the balance. Re-pricing a menu shows up next week. Renegotiating a lease shows up next quarter. Automating a process shows up after you build it. Score each lever fast (days), medium (weeks), or slow (a quarter or more). Axis two: reversibility. How easily you can undo it if you are wrong. Re-pricing is reversible. Cutting a dead shift is reversible. Signing a new long lease, making a role redundant, or gutting the product is not. Score each lever reversible, semi-reversible, or one-way. Now sequence: The rule the grid encodes: pull fast and reversible first, and never make a one-way, slow-to-pay move under pressure. Fast-and-reversible levers buy you the time to think. They are cheap to be wrong about. The one-way moves, closing a site, making people redundant, signing a long commitment, are the ones you must not improvise, because they cannot be walked back and they rarely pay fast enough to be worth the risk of doing them scared. Write the trigger for the priority-4 moves in advance. Something like: if the rebuilt margin stays below the line for two consecutive months after the fast levers are pulled, then execute the contraction plan already drafted. Deciding the rule with a clear head, then acting on it without re-litigating when frightened, is the entire value of the grid.
Section 3
Worked example: the UK employer National Insurance shock
Watch the worksheet run on a real shock. In April 2025 the UK changed employer National Insurance. The rate rose from 13.8 percent to 15 percent, and the threshold where employers start paying dropped from 9,100 pounds a year to 5,000 pounds (Deloitte TaxScape on the Autumn Budget 2024 measure; Xero UK's summary). Stacked on top: the National Living Wage rose 6.7 percent to 12.21 pounds an hour (GOV.UK), and business rates relief for hospitality was cut from 75 percent to 40 percent (GOV.UK guidance on the 2025/26 scheme). By June 2026, 23 percent of pubs and restaurants were operating at a loss, up from 15 percent three months earlier (UKHospitality with the British Beer and Pub Association and partners, reported via AOL). An operator who trimmed would have shaved a few shifts and hoped. An operator who rebuilt would have caught the shape change. Take a worker on 20,000 pounds. Under the old rules the employer paid NI on earnings above 9,100 pounds, so 13.8 percent of 10,900, which is 1,504 pounds. Under the new rules it is 15 percent of 15,000, which is 2,250 pounds, a 50 percent jump for the same person doing the same job (The Access Group's hospitality analysis). Worse, a part-timer on 6,000 pounds who used to generate zero employer NI now generates 150 pounds, because the threshold dropped below their pay. Enter that per-head cost in step three and the rebuilt contribution margin on a covered table falls in a way that trimming would never have surfaced, because the cost stopped scaling with wages and started applying per head. Now the lever grid. Fast and reversible: re-price the items customers buy on habit, kill the genuinely dead trading hours, and claim the Employment Allowance, which rose to 10,500 pounds in April 2025 with the old eligibility cap removed, so a large share of smaller employers pay no employer NI at all (The Access Group). Those hit the bank inside a fortnight and can be undone. Slow and reversible, started now: cross-train staff so one person covers two roles in a quiet hour, and shorten supplier commitments as they come up for renewal. One-way and slow, on a trigger only: closing a site or restructuring headcount, executed only if the rebuilt margin stays underwater after the fast levers, never on the first bad week.
Section 4
Why two models, lightly, sit under this
The worksheet is not arbitrary. It rests on two disciplines worth naming, briefly, because knowing why the tool is shaped this way keeps you from misusing it. The rebuild is comparative statics: move one variable, hold the rest still, and trace exactly where the equilibrium lands. Its whole job is to force you off the phrase "costs are up" and onto the specific line that moved and the specific way it now behaves. Its limit, stated honestly, is that the world moves back. Customers, staff and competitors all react at once, so the clean rebuilt number is the opening position, not the final one. Re-run it as reality responds. The lever grid is real-options thinking: the right to wait has value, and a reversible move preserves that right while a one-way move spends it. That is why fast-and-reversible levers come first and one-way levers wait for a trigger. Its limit is that some doors close on their own. Wait too long on a genuinely necessary contraction and the option expires whether you exercised it or not. The trigger exists to stop the option-value logic from curdling into paralysis.
Section 5
What the playbook cannot see
The rebuild prices your business absorbing a shock. It does not price the next shock, or politics itself. Whether the relief gets extended, whether the following budget adds or removes a cost, whether your sector gets singled out again, none of that is visible in your unit economics, because it comes from a system that does not read your cash flow. The honest posture is that you are building for robustness, not running a forecast. You cannot rebuild your way to certainty about the next decision. You can only rebuild fast enough that no single decision catches you still modelling the old world. The worksheet also assumes you have levers to pull. If the rebuilt margin cannot reach breakeven at any realistic volume on the new cost function, no operating lever fixes that. That is a financing or exit decision wearing an operations costume, and the rebuild's real service is that it tells you which of the two situations you are in before you spend three months pulling levers that were never going to close the gap.
Section 6
The fitness test
You are ready for the next policy shock if you can, inside a week of it landing, rebuild your contribution margin from a blank page on the new rules, name the single cost line whose shape changed, and place your available levers on the speed-to-cash and reversibility grid with the one-way moves parked behind a written trigger. If you can do that, a shock is a bad quarter you navigate on real coordinates. You are not ready if your response to a shock is to open last year's budget and trim, because that is optimising a business that stopped existing the day the rule changed. The operators who survive the next decade of policy volatility will not be the ones who trimmed hardest. They will be the ones who rebuilt fastest, on the new shape, before their bank balance did the rebuild for them.