Google Ads Performance Planner 2026: Q4 Budget Guide

Every August the same conversation happens: finance wants a Q4 number, and the only justification anyone has is “we spent about this much last year.” The Google Ads Performance Planner exists to answer exactly that question — it models the auction, draws a spend-to-conversions curve, and shows you where the next dollar still pays for itself. The problem is that most advertisers treat it as a fortune teller: open it, see “+38% conversions,” feel good, and never inspect a single assumption behind the number.

This guide covers what the planner actually computes, what Google quietly removed from it in 2026, why half your campaigns don’t show up in the list, how to read the inflection point of the curve, and how to build a Q4 plan in twenty minutes that survives contact with November.

What Performance Planner actually models

Performance Planner is an auction simulator, not a spreadsheet with percentages. Google takes the last 7–10 days of your campaign data (up to 30 in some campaign types), runs it against its model of the search auction, and simulates dozens of scenarios: what happens to conversion volume at 10%, 30%, 80% more budget, at a lower target CPA, with an additional campaign in the mix.

The output is a curve. Budget on the X axis, conversions on the Y axis. The valuable part is not the endpoint — it’s the shape. Early on the curve climbs nearly linearly: each additional dollar buys roughly the same number of conversions. Then it flattens. That bend is the point of diminishing returns, and it answers the only budget question that matters: how much can we push before the economics stop working?

Three caveats Google prints in small type and everyone skips:

The forecast assumes everything else stays constant — same creative, same landing pages, same competitive set, same demand. In Q4, none of those hold.
The simulation knows auction seasonality from history. It does not know about your promotion, your stockout, or the well-funded competitor who entered your category in October.
The model leans optimistic on budget increases, because it was trained on auction data where more budget almost always buys something. The real question is the price of that something.

So rule number one: the planner is a hypothesis about the shape of your return curve, not a promise of outcomes. Trust it on where the curve bends. Don’t trust its absolute conversion counts.

What changed in the Google Ads Performance Planner in 2026

Two changes landed with almost no fanfare, and both break workflows people had been using for years.

On March 9, 2026, Google removed Display and Video campaign support from Performance Planner, along with any plan built around impression share metrics (impression share, top IS, absolute top IS). Existing plans containing those elements stopped opening entirely — you can no longer view, edit, or export them. If you had a “master annual plan” with a display line in it, that file is gone.

What still works:

Campaign type	Supported	Notes
Search	Yes	Primary use case, most accurate simulation
Shopping (Standard)	Yes	Smart Shopping already migrated into PMax
Performance Max	Yes	Coarser forecast — the internals are a black box
Demand Gen	Yes	Added later, needs meaningful volume
App	Yes	Separate logic depending on conversion type
Local	Yes	Limited, depends on geo data density
Display	No (since 03/09/2026)	No native in-platform planning left
Video / YouTube	No (since 03/09/2026)	Reach Planner remains for reach goals
Impression-share plans	No (since 03/09/2026)	Model these manually from auction data

The second change is the partial replacement: in January 2026 Google launched cross-channel budgeting inside its analytics stack, with independent projections and scenario planning. In effect, upper-funnel planning has been moved out of Google Ads. For display and video that’s now the only native option, and it models channel contribution rather than the auction itself — a different tool with a different kind of accuracy.

Practical consequence: if you run upper-funnel media, planning now splits into two processes. Performance sits in the planner, awareness sits in the analytics layer, and the link between them is something you have to measure yourself — usually with geo experiments and incrementality testing, because no planner will tell you how much display lifted your branded search.

Why your campaigns don’t appear in the plan

The most common complaint is that half the account is missing from the campaign selector. That’s by design. Each campaign type has a data floor below which simulation is meaningless. Approximate requirements — Google adjusts these without announcements, so verify in Help at the time you plan:

the campaign has been running and spending for at least 7–10 consecutive days without pauses;
it accumulated roughly 3,000+ impressions over the past week in search;
it has at least ~10 conversions in the period for conversion-based plans — below that the curve is drawn from noise;
it uses a supported bid strategy: tCPA, tROAS, Maximize Conversions, Maximize Conversion Value. Manual bidding without a conversion goal isn’t plannable;
it isn’t constrained by hard limits like a campaign total budget or an imminent end date.

Treat those numbers as an order of magnitude, not a contract. The underlying logic is stable: no volume, no forecast. A campaign that can’t qualify for the planner is usually also too small to base a budget decision on.

One trap deserves its own mention: recent structural change. If you rewrote half the keyword set, switched bid strategy, or relaunched the campaign last week, the history is contaminated — the planner will model a campaign that no longer exists. Give it two to three weeks to stabilize. The corollary matters more: don’t schedule restructures right before the season when you need the forecast. Sequencing that properly is covered in the guide to seasonality and budget planning.

Building a plan in twenty minutes

Tools and settings → Planning → Performance Planner. The “+” button creates a new plan.
Pick the period. For Q4, build three monthly plans rather than one quarterly plan. October, November, and December have genuinely different auctions, and a quarterly average hides the November spike.
Select campaigns deliberately. Don’t throw everything into one plan. Group by budget logic — brand separately, non-brand search separately, PMax separately. Otherwise the planner will happily shift money between things with incomparable economics.
Choose the key metric. Conversions, conversion value, clicks, or impressions. For performance work it’s conversion value if you pass it, conversions if you don’t.
Set the anchor. Two modes: “here’s my budget, show me the outcome” and “here’s my target outcome, show me the budget.” The second is the honest one when you’re defending a number internally.
Work the curve. Hover across points — the planner shows incremental conversions and incremental CPA at each. This screen, not the summary at the top, is where the decision lives.
Adjust per campaign. Inside the plan you can move budget between campaigns and watch the total shift. The planner will propose an “optimal” allocation — optimal to its model, not to your margin structure.
Export and archive the snapshot. A month later you compare forecast to actuals. After three such comparisons you’ll know your account’s systematic deviation from the model — and that number is worth more than the forecast itself.

Reading the curve: incremental CPA beats average CPA

Beginners read the endpoint: “at $30k we get 900 conversions.” Experienced buyers read the derivative: what does each additional conversion cost?

Incremental CPA = (Budget₂ − Budget₁) ÷ (Conversions₂ − Conversions₁)

Worked example:

Budget, $	Conversions	Average CPA, $	Incremental CPA of the step, $
20,000	640	31.3	—
24,000	730	32.9	44.4
28,000	790	35.4	66.7
32,000	825	38.8	114.3

Average CPA drifts gently from $31 to $39, which looks survivable on a slide. Incremental CPA goes from $44 to $114 — the last $4,000 buys conversions at nearly three times the price of the first. If your allowable CPA is $50, the correct read of this plan is roughly $24,000, not $32,000, and the remaining budget belongs in a different source of volume.

None of this works without your own allowable CPA, derived from margin and repeat purchase rather than from habit. That derivation is covered in the breakdown of media buying unit economics and max CPA from LTV.

The second read: when the curve is already flat

If the curve flattens almost immediately, you’ve already bought the available volume and additional budget is just bidding up the same auctions. Confirm it outside the planner: check lost impression share due to budget versus due to rank. Budget-related loss near zero with high rank-related loss means money is not the constraint — relevance, assets, and landing experience are. How to interpret those columns is covered in the piece on impression share and auction insights.

What the planner can and cannot do

Can	Cannot
Show the shape of your return curve	Account for a new offer or discount
Compare scenarios across tCPA/tROAS values	Predict a new competitor entering
Reallocate budget between campaigns	Distinguish margin between products
Reflect historical auction seasonality	Know about your own promotion calendar
Give a defensible incremental CPA	Guarantee absolute conversion counts
Export to a spreadsheet	Plan Display, Video, or impression share (since 03/09/2026)

Seven mistakes that turn a plan into wishful thinking

Planning on dirty conversions. If contact-page views and phone clicks are counted alongside purchases, the planner optimizes noise. Clean the conversion model first — ideally with proper first-party signal in place.
Mixing brand and non-brand in one plan. Brand delivers cheap conversions that will absorb the “optimal” allocation. Plan them separately.
Reading the summary instead of the curve. “+38% conversions for +52% budget” is negative incremental economics dressed up as good news.
Planning right after a restructure. The model is describing a campaign that no longer exists.
Applying the plan with one click. The planner can push recommended budgets and targets in bulk. That resets bid strategy learning across several campaigns simultaneously.
Ignoring conversion lag. If deals close over 14 days, a forecast built on last week’s data systematically understates conversions — and you underinvest in December.
Not archiving the forecast. Without forecast-versus-actual comparison you never learn your own correction factor, and every account has one.

A twelve-week runway into Q4

Window	What you do	Why then
Weeks 1–2 (late August)	Clean the conversion model, verify tags, finish the negative list	Forecasts are only as good as the inputs
Weeks 3–4 (September)	Freeze structure — no migrations past this point	You need two to three weeks of clean history
Week 5	Build three plans: October, November, December	Different auctions, different curves
Week 6	Compute incremental CPA per step, cut the plan at your threshold	Defend the budget with arithmetic, not faith
Weeks 7–8	Run an experiment on the contested hypothesis	Test before the season, not during
Weeks 9–10	Raise budgets in 15–20% steps	A single large jump resets learning
Weeks 11–12	Rebuild the plan on early-October actuals	A forecast is a living document, not a PDF

The experiment step isn’t filler. Any contested change the plan implies is far cheaper to validate with a controlled split test than to discover on full traffic in November — the mechanics are covered in the guide to experiments in Google Ads and Experiment Center.

Planner versus Recommendations: two opposite tools

People conflate them constantly, and their logic runs in opposite directions. The planner shows you a curve with the price of each next step — you decide where to stop. Recommendations show you a list of actions with promised uplift, and most of them expand spend by construction. One gives you leverage; the other asks for it. Which recommendations are worth applying and which belong in the bin is covered separately in the piece on optimization score and Google Ads recommendations.

When not to trust the planner at all

Long sales cycles (B2B, real estate, education). The model works on short windows; your real result shows up months later.
Demand shocks. A product launch, a viral news cycle, a regulatory change — there is simply no history to simulate against.
An account after a suspension or migration. Until new history accumulates, forecasts are fiction. What matters more there is how fast you can restore volume — agency Google Ads accounts let you avoid rebuilding history from zero.
Under ten conversions a week. The task isn’t planning, it’s building signal volume: micro-conversions, enhanced conversions, consolidated goals.

Performance Planner does not answer “how much will we make.” It answers “where does the next dollar stop paying for itself.” It doesn’t know the first. It knows the second better than you do — if your data is clean.

Checklist before you present the plan

The conversions in the plan are revenue events, not micro-actions.
Brand and non-brand are modeled separately.
Incremental CPA is written out for every budget step.
The plan is cut at your allowable CPA, not at the prettiest number.
Conversion lag is accounted for in the December line.
The forecast is archived for later comparison against actuals.
Contested changes are queued as experiments, not applied on the spot.

If account infrastructure is its own bottleneck — spend limits, thin history, access to the geos you need — fix that before planning rather than after. That’s the layer PPC Rebels services and PPC Rebels agency ad accounts cover: a stable platform on which a forecast is worth building in the first place.

FAQ: Performance Planner questions

Is Performance Planner free?

Yes, it’s built into Google Ads at no cost. The only gate is statistical: campaigns without enough recent data won’t appear as plannable.

Why is my campaign missing from the plan builder?

Usually insufficient recent volume (impressions or conversions), an unsupported campaign type, an unsupported bid strategy, or a recent relaunch. Since March 9, 2026, Display and Video campaigns aren’t supported at all.

How accurate is the forecast?

The shape of the curve is usually reasonable; the absolute numbers are not. A 15–30% gap between forecast and actuals is normal in practice, and it tends to be systematic per account. Archive forecasts, compare with actuals, derive your own correction factor.

What happened to Display and Video planning?

Google removed both on March 9, 2026, together with impression-share-based plans, and existing plans containing them stopped opening. Reach Planner still covers reach goals, and the cross-channel budgeting released in January 2026 is the new scenario-planning layer for non-search media.

Should I use the “apply plan” button?

Technically you can; practically, don’t. Changing budgets and targets across several campaigns at once restarts bid strategy learning everywhere. Step budgets up 15–20% at a time and watch the learning period.

Does the planner account for seasonality?

It accounts for auction seasonality drawn from history — the general demand pattern of your category. It does not know about your own promotion, inventory constraint, or a competitor’s campaign. Add those manually.

How often should I rebuild the plan?

Monthly in stable categories. Every two weeks heading into a season, and weekly once the peak starts. A forecast nobody revisits after the first week of actuals is just a document.

Does Performance Planner work with Performance Max?

Yes, PMax is supported, but its forecast is coarser than search. The campaign blends surfaces and formats, so the auction simulation is less direct. Read the PMax curve as an order-of-magnitude guide.

What if the curve is almost flat?

You’ve bought the available volume. Check lost impression share: low budget-related loss with high rank-related loss means additional spend won’t help — relevance, assets, and landing experience will.

How is this different from Keyword Planner?

Keyword Planner forecasts volume and cost for keywords when building new campaigns — it’s a research tool. Performance Planner works on live campaigns with real history and answers a budget question, not a keyword question.

Is the planner useful if my budget is fixed?

Very. With a fixed total you use it for reallocation: it will show which campaign delivers more conversions for the same money if you shift budget out of a more expensive one.

Google Ads Performance Planner 2026: Build a Q4 Budget You Can Defend