Cutting the budget by 20%: preserving the system, not only the average return.
Evidence status
What was reproduced, and what it does not prove.
The CSV and the Python run verify the three plan comparisons, the stress constraint and the robust choice in the illustrative scenario. They predict no market performance.
- Technical specification of the 20% budget cut, specification in French (opens in a new tab)
v1, 31 August 2026 · CC BY 4.0 · specification in French · Method document written by marketing-mix.net. The figures are illustrative: they are not market measurements.
Key distinctions
Three conditions before calculating.
Structural steps
A large cut can close a country, a team, a contract or a testing capability. Those discontinuities must be modelled as steps, not as a smooth curve.
Joint stress
Returns, competition and costs can deteriorate together. Testing each uncertainty separately understates the possible loss when several unfavourable assumptions occur at once.
Cost of rebuilding
Reactivating a campaign is quick; rebuilding a database, a local presence or a creative capability can take months. That delay and its cost belong in the reversibility calculation.
Method
Moving from a marginal cut to a resilience plan
- 01Map the steps
Split the spend into tranches, then identify floors, contracts, dependencies, teams and exit costs; flag the cuts that destroy an entire capability.
- 02Build the scenarios
Estimate the contribution loss in a central scenario, a low-efficiency scenario and a competitive-reaction scenario, and add reactivation delays and costs.
- 03Trace the frontier
Compare the plans by savings achieved, expected loss, maximum loss and risk concentration; discard those that exceed the tolerance, even when they dominate at the centre.
- 04Sequence the reduction
Start with the reversible tranches, measure the leading signals, then cross the steps only after review, and keep a correction reserve.
The expected loss × stress loss frontier
Three plans reach the same €200k saving but expose the company to very different losses when the assumptions deteriorate.
The robust plan costs €16k at the centre but avoids €52k under stress
Three plans each save €200k out of a €1m budget. The concentrated cut loses €88k in the central scenario and €180k under stress. The robust cut loses €104k at the centre and €128k under stress. The pro-rata cut loses €120k and €150k respectively.
Additional central cost of the robust plan against the concentrated one: 104 − 88 = €16k. Protection under stress: 180 − 128 = €52k. With a maximum loss tolerance of €140k, only the robust plan satisfies the constraint.
The robust plan is retained. It gives up a small part of the central optimum to preserve capabilities and to reduce the loss sharply when the assumptions deteriorate.
Acceptance conditions
What must be true to act.
- 01
The cuts that cross a structural step are identified.
- 02
The loss is calculated in at least one joint stress scenario.
- 03
The cost and the delay of rebuilding are included.
- 04
The maximum loss of the plan stays under the approved tolerance.
Limits
What this analysis does not prove.
- Stress scenarios do not cover every possible shock.
- Competitive reaction remains hard to quantify.
- A sequenced reduction can be incompatible with a cash-flow emergency.
References
Works cited.
- Ben-Tal, A., El Ghaoui, L. & Nemirovski, A. (2009), Robust Optimization.
- Bertsimas, D. & Sim, M. (2004), “The Price of Robustness”, Operations Research, 52(1), 35–53.
- Keeney, R. L. & Raiffa, H. (1976), Decisions with Multiple Objectives.