Reallocation After a Budget Cut
"The client cuts 30% - where do we take it out?" Spreading a cut evenly across all channels is the obvious reflex - and rarely the best path. On the model you can show where a cut costs the least effect: where a channel is already working at its saturation limit, rather than where it carries other channels via the halo effect.
Why the across-the-board cut is weaker
The lawnmower hits the wrong channels
Cutting every channel by the same percentage treats them all the same - though they aren't. A channel near saturation loses barely any effect from a cut; a channel in the steep part of its curve loses disproportionately. The across-the-board cut takes the same amount from both and thus often hits exactly the wrong one.
What the model makes visible
The saturation model shows for each channel where it sits on the curve. The halo effect shows which channel pulls others along. Together that gives an order: cut first where the marginal return is already flat - last where a channel carries the mix.
How the reallocation works
- Load the starting plan. Bring the current state - the master strategy or the last plan - into the model as the starting point.
- Redistribute the reduced budget. Set the monthly budgets of the individual channels anew by hand - this is where the actual reallocation happens. Don't take the same amount everywhere, but specifically where it costs the least.
- Read saturation and halo. From the saturation model, spot which channels work near their limit (α) and lose little when cut; from the halo, see which channels carry others and therefore should be spared.
- Check the effect at once. The result boxes update with every input - each shift immediately shows what happens to customer count, break-even and net profit.
- Compare variants and save. Use "compare plans" to set the reallocated variant against the starting plan, pick the most viable one and save it as a plan or JSON backup.
Budget cut or market downturn?
Two things that are easily confused. When the client cuts their budget, you work on the channel budgets - the strategy stays, just with fewer means. A market downturn or a political tailwind is something else: a lockdown pushes results down, a new law - a sector subsidy program, say - lifts them, without anything changing in the underlying strategy. That is exactly what the global change in the base data is for: a percentage that raises or lowers all channel results at once.
When a customer target still stands
If a certain customer count still has to be reached after the cut, the target back-calculation takes over: set the target customer count, and the model works back the budget needed for it - along with a table of how it distributes across the channels. Saturation, seasonality and time lags are included, so the scaling factor is not linear.
Play through a cut on the model and find the gentlest reallocation: try the tool
Talking points for the conversation
- "Don't cut everywhere equally - cut where it costs the least." Moves the conversation from how much to cut to where.
- Saturation as an argument: a channel at its limit gives up budget without losing many customers - that's the first address.
- Halo as a protective argument: a channel that carries others (offline to SEO, say) costs twice when cut - once directly, once at the recipients.
- Put two variants side by side - the harder and the cautious cut - and let the client decide. Consulting instead of dictating.
Common thinking traps
- Cutting across the board and calling it reallocation. Lowering every channel by the same percentage is a shrink, not a redistribution. The actual reallocation only emerges at the individual channel budgets - and the global percentage slider in the base data isn't meant for it anyway; it models external events.
- Reading saturation and halo as exact values. Both rest on estimated curve parameters (α, lag, decay), not on measured causality. The model shows the direction "cutting costs less here", not a guaranteed number.
- Looking only at the direct channel contribution. Whoever ignores the halo cuts a channel away seemingly cheaply and unwittingly loses the effect it fed to other channels.
Frequently asked questions about reallocation
Why not just cut every channel by the same percentage?
Because channels react differently to a cut. A channel near its saturation limit loses barely any effect, a channel in the steep part of the curve loses disproportionately. The across-the-board cut hits both the same - and thus often the wrong one.
What is the global change in the base data for?
For external events that affect all channels at once without changing the strategy itself - a lockdown or a recession pushes results down, a new law or subsidy program lifts them. A client's budget cut is something else: it is mapped at the individual channel budgets, not through this global slider.
How do I tell which channel can give up budget?
From the saturation model. If a channel works near its limit (α), extra budget there brings barely any effect anyway - and a cut costs correspondingly little. The halo effect, conversely, shows which channels carry others and should be spared.
Do I see the effect of a reallocation right away?
Yes. The result boxes update with every input - each budget shift immediately shows what happens to customer count, break-even and net profit. That way a variant can be played through in the conversation.
What if a customer target remains despite the cut?
Then the target back-calculation helps: you set the target customer count, and the model works back the budget needed for it and shows in a table how it distributes across the channels. Saturation, seasonality and lags are included, so the scaling factor is not linear.
