In short: Performance Planner simulates spend → result with real auction data, and answers the director's question: "if we invest 30 % more, what do we get?". What you need to read isn't the total but the slope: how much the next euro yields. It doesn't see your margin or your quality and assumes nothing changes, so it's for direction and order of magnitude.
Tools → Planning → Performance Planner. It is Google's curve calculator: using your campaigns' auction data it simulates how many conversions (or clicks, or value) you would get at different budgets and targets over the coming period. Used well, it answers the boss's question: "if we invest 30% more, what do we get?" Used badly, it is a promise nobody will be able to keep.
What does Performance Planner do?
- You pick campaigns (Search, Shopping, PMax, Demand Gen, app; since March 2026 not Display or video, and no plans with impression share metrics).
- You define the period (next month, quarter) and a key metric (conversions, value, clicks).
- Google simulates using its auction data, seasonality and your history; it draws the spend → result curve per campaign and for the plan as a whole.
- You can fix a total budget or a target (CPA/ROAS) and the Planner splits it across campaigns to maximise the result.
- It shows, per campaign, the suggested budget/target change and the forecast result; the plan can be applied with one click (better to export it and apply by hand, in steps).
- It also lets you test changes to bids or budget targets before applying them, to see their forecast effect on your existing campaigns.
The forecasts refresh daily using the last 7-10 days, adjusted for seasonality.
How do you read Performance Planner?
- The plan's curve: the current point and the proposed point; the slope between them is the marginal CPA (lesson 2) that Google is estimating.
- Per campaign: the proposed split tells you where Google thinks the next euro performs best — usually where there is lost IS to budget with a good CPA.
- Scenario comparison: build two or three plans (current budget, +20%, +50%) and compare the forecast results. The difference between +20% and +50% is usually the flat zone becoming visible.
What are the Planner's limits?
- It is a Google forecast built on Google data: it cannot see your margin, your capacity or the quality of your leads. A plan that "maximises conversions" maximises whatever you count as a conversion.
- It extrapolates: the further from current spend, the less reliable. +20% is reasonable; ×3 is fiction.
- It assumes nothing else changes: competition, creatives, website.
- Approximate seasonality: fine for a normal month; for sales periods or Christmas, check it against your own previous year.
- New campaigns or thin data: unreliable forecasts.
Rule: the Planner gives you direction and order of magnitude; reality supplies the exact figure.
What is the Planner actually good for?
- Negotiating budget: a +X% scenario with forecast conversions and a marginal CPA is an argument backed by data, with the caveat that it is an estimate.
- Splitting a fixed budget across campaigns: its proposal is a good starting point that you then adjust with the business-goal judge (module 2).
- Spotting the flat zone: campaigns whose forecast does not improve with more money.
- Planning the month: daily budgets per campaign that add up to a coherent monthly total (lesson 5).
What is the Planner not good for?
- Applying the whole plan in one go: it resets learning and mixes changes together.
- Treating the forecast as a commitment to management.
- Planning Display/video (no longer supported) or accounts with dirty measurement (rubbish in, rubbish forecast).
💡 Ninja trick: the Planner is a monthly snapshot; the curve moves every week. Budget allocation in the Suite is not driven by a one-month forecast but by small daily steps guided by target attainment, lost IS and CPA trend. Use the Planner to decide the month's total and to negotiate it; leave the fine-grained allocation to the short cycle.
What you should remember
- It simulates spend → result from auction data; Search, Shopping, PMax, Demand Gen (no Display/video since 2026).
- Read the slope (the marginal) and the proposed split; compare scenarios.
- Limits: it cannot see margin or quality; it extrapolates; it assumes nothing changes.
- Use it for direction, order of magnitude and negotiation; apply in steps.