Platform Mix
“What’s the optimal split across Google / Meta / Microsoft?” Curves fitted per campaign over 91 days, aggregated to platform level, then solved for the split where marginal return is highest.
[05/07]Budget Planner
/ the product tour
Budgets planned on curves, not hunches.
Given a total budget, what’s the optimal split across Google, Meta and Microsoft — or across whole portfolios? Budget Planner answers both with the same response curves the live optimiser runs on, so the plan and the execution never disagree.
[01]The modes
/ two planning questions
Two questions every budget meeting asks.
One switcher, two modes — both built on the saturation curves fitted to your own campaign history, not industry benchmarks.
Portfolio Mix
“Given a total budget across portfolios, where does each increment go?” Pick an objective, choose portfolios, set the total — each budget increment flows to whichever portfolio has the highest marginal return at its current level.
The same curves that execute
The planner isn’t a separate model that drifts from reality — it reads the exact response curves the nightly optimiser allocates with. What you model is what the engine will do.
Honest about confidence. Every curve carries a 95% confidence
interval from bootstrap resampling. Narrow bands mean a confident split; wide bands
flag low-confidence curves — and the live allocator automatically dials back
how hard it leans on them to match.
[02]The output
/ from plan to action
See the gain. Then apply it.
A plan you can’t interrogate is a guess with a chart. Every model run shows its working — and applies through the same staged, reviewed pipeline as everything else.
Efficiency curve
Current vs optimised split across budget levels, in three views — Revenue, ROAS, and Marginal ROAS. The marginal view shows exactly where each unit of spend hits diminishing returns.
Revenue uplift waterfall
Per-portfolio revenue change from current to optimised allocation — so you can see precisely where the gain comes from before you commit to it.
Spend ceiling
Each portfolio’s addressable ceiling is estimated from impression-share data. The planner won’t model spend a portfolio can’t actually absorb.
Ramp, don’t lurch. Apply the new split immediately, or
Step Over N Days — the nightly scheduler walks budgets to
the target one step per night, with the same staging, review and read-back as
every other change.