Performance Max is testing channel sliders. Read the channel report at the margin before you move one

KEY TAKEAWAYS

1. The slider is a price, not a budget. It changes the CPA Performance Max will accept on one channel.
2. The channel report shows average CPA. The slider acts on the next conversion, which usually costs more than the average.
3. Whether the campaign is limited by budget decides where freed money goes. Without a budget limit, it is simply not spent.
4. Confirm any move with a holdout. Google's Conversion Lift needs 1,000 conversions and a $5,000 budget.

Some Google Ads accounts now show a Channels setting inside Performance Max, marked Alpha. It looks like a budget dial. It is closer to a price tag. And the report most people will use to set it, the channel performance report, shows the wrong price.

What changed in Performance Max between August and October?

Three things, close together.

• 17 August 2026. Budget-limited campaigns on Target CPA or Target ROAS, Performance Max included, began to deliver closer to their target instead of beating it. Google's help page says multi-channel campaigns may see traffic shift across channels as a result.
• 25 August 2026. Heidi Sturrock reported a Channels setting in alpha. Search Engine Land covered it that day and Search Engine Roundtable the next. Each channel gets a slider running from lower prioritisation, through Auto, to higher prioritisation.
• 7 October 2026. Google's Conversion Lift setup page was rewritten to list Search and Performance Max as supported campaign types.

A separate pilot, first spotted in April and confirmed by Google to Digiday in July as limited to a small group of advertisers, adds two checkboxes that remove Search partners and the Display Network outright.

Google has not announced the slider. A search on 13 October found no report of a wider rollout.

What does the slider actually change?

It changes the cost per conversion the campaign will accept on one channel. Search Engine Land, Search Engine Roundtable, TechWyse and Veritas Digital all describe it the same way: a higher setting relaxes that channel's CPA, a lower one tightens it. The in-product text tells advertisers to use it if they "value a channel more or less based on its importance to your business."

Three things it is not:

1. Not a budget split. Spend still follows the auctions.
2. Not an off switch. Veritas notes a channel at a lower setting keeps serving wherever it finds conversions inside the tighter bar.
3. Not documented. MetricFixer found no public API field, and nobody has published how big each step is or how it behaves under Target ROAS.

One source disagrees on direction. PPC News Feed wrote that lowering a channel's priority makes Google raise the CPA bid there. That cannot fit the in-product wording, since valuing a channel less cannot mean paying more for it. Every other report we read agrees on the opposite direction, and we follow them.

Brainlabs compares the slider to conversion value rules. Under a CPA target the two ideas are the same: count a channel's conversions as worth more, and the bidder will pay more for them. Google's own value-rules page adds a useful warning. Smart Bidding already sees the segments that convert better, so a rule only helps when the value it adds is something the conversion data does not show. The same test applies here.

Which control does which job?

Google's Demand Gen page is candid about the last row. Its bidding uses marginal cost optimisation, so a channel can serve less than you expect. To force spend onto a channel you isolate it with a higher CPA, and Google warns that can be less efficient. That is the trade the PMax slider now offers inside one campaign. For the view-through settings that sit alongside it, see our Demand Gen view-through guide.

Why is the channel report the wrong input on its own?

Because it shows averages, and the slider acts at the margin.

Google says so itself. Its guide to evaluating Performance Max defines marginal ROI as the extra return from extra spend. It adds that a channel with a lower average return can still have the highest marginal return in a single auction. The channel report help page says much the same: a channel with no spend and no diagnostic issue usually means the campaign found better returns elsewhere at that moment.

Attribution adds error in both directions.

• Upper-funnel channels can be under-credited. Veritas, TechWyse and Brainlabs all give the same example: someone sees a YouTube ad and converts later through Search.
• Search can be over-credited. Optmyzr found Performance Max serving on search terms that matched Search keywords in 91.45% of the accounts it studied (February 2025 data). Some conversions credited to Search inside PMax may have come through your Search campaigns instead.
• View-based conversions sit in the same table. The report's ad event type segment includes Engaged-View and Impression as well as clicks. Split them before you compare channels.

Benchmarks do not settle it either. TNT Growth's benchmark of 53 accounts over 365 days found Search took a median 61% of PMax spend, but individual accounts ranged from 2% to 99%. YouTube's median was 10%, Display and Search partners 3% each, and Gmail, Discover and Maps 0%. Optmyzr's holiday data from more than 4,200 accounts shows YouTube with nearly 10% of PMax impressions but about 2% of conversion value. Both are attributed figures, not lift. A spread that wide says your mix is your own; it does not say which way to move it.

Does it matter whether the campaign is limited by budget?

Yes, more than anything else on this page. This is our reading of Google's documentation, not a statement Google has made about the slider.

• Limited by budget. Since 17 August the campaign spends toward its target. Tighten one channel and the budget it no longer spends there flows to other channels, at their next-conversion cost.
• Not limited by budget. Google's FAQ says unconstrained campaigns scale in line with the target. Tighten one channel and that money is simply not spent. Nothing is reallocated. You save cost and lose conversions.

So the same slider move is a reallocation in one campaign and a cut in another. Check the status column before you touch it.

When does lowering a channel pay?

Here is a break-even rule. It is our own calculation, not a Google formula.

Moving budget B from channel C to channel R pays only if:

B ÷ marginal CPA of R > B ÷ marginal CPA of C + A

A is the number of conversions elsewhere that depended on channel C and now disappear. No channel report shows A. Only a holdout can estimate it.

Hypothetical example. A budget-limited campaign has a $50 target CPA and spends $10,000 a month for 200 conversions. The channel report shows:

On averages, moving $1,500 from YouTube to Search looks like swapping 15 conversions for 37.5. Now assume the margins. The $1,500 that a lower YouTube setting removes was its least efficient spend, at $150 a conversion, so 10 are lost. The extra Search spend buys conversions at $75, not $40, so 20 are gained. The real gain is 10, not 22.5. And if YouTube helped bring in more than 10 of the Search conversions, the move loses.

Without a budget limit, the same move saves $1,500 and loses 10 conversions plus A. It pays only if a $150 conversion was already worth less than $150 to you.

You can estimate a marginal CPA without new tools. Divide the change in campaign cost by the change in campaign conversions between a baseline period and the test period.

What should you do in each situation?

How to test a slider change in seven steps

1. Record the starting point. Note which channels your account groups together (Spilno saw Gmail grouped with Display, others see them separate), the target, the budget and whether the campaign is limited by budget.
2. Pull 90 days of the channel report. Segment by ad event type and set view-based conversions apart.
3. Compare each channel with target over several windows, such as 30, 60 and 90 days. Act only on a gap that holds in all of them.
4. Write down why the channel is worth more or less than its conversions show: margin, offline value, new customers, or a lift result. No reason means leave it on Auto.
5. Change one channel, one step, in one campaign. Keep budget, target and assets fixed. Google lists no Performance Max experiment type for this setting, so use an unchanged, comparable campaign as your reference.
6. Wait two conversion cycles, then compare campaign totals. Google's channel report page asks for one to two weeks after significant changes. Work out the marginal CPA from the change in cost and conversions.
7. Confirm with a holdout. Conversion Lift needs 1,000 observed conversions, a $5,000 budget, a holdback between 1% and 50% and, Google recommends, more than 14 days, aiming for 90% power. Google's comparison page still routes Performance Max user-based lift through a representative, so ask. Below that bar, a geo test is the alternative; our geo test guide covers how to choose one.

What this means for the rest of the account

The slider moves cost between channels inside one campaign. It does nothing about overlap with your standalone Search campaigns, which the AI Max upgrade has made broader. Treat Performance Max, AI Max and brand Search as one budget question, measured with one holdout where you can.

Paid and organic now feed the same brand searches, and increasingly the same AI answers. Ethos Engine is building tools to measure brand visibility across both. If that is useful, join the waitlist.

Questions

Does the Performance Max channel slider set a budget for each channel?

No. Every report of the alpha describes it as a change to the cost per conversion the campaign will accept on that channel. Higher prioritisation lets Performance Max pay more there. Lower prioritisation makes it pay less. Spend follows from that, auction by auction, so no slider position guarantees a share of budget or a minimum volume.

Should I lower YouTube in Performance Max if its CPA is above my target?

Not on the average alone. Google's own guidance says a channel with a weak average return can still have the best marginal return in a given auction. Check whether the campaign is limited by budget, split view-based conversions from clicks, and estimate what the spend you cut was really buying before you move anything.

Who can use the Performance Max channel slider?

Only advertisers in an alpha test. It was first reported on 25 August 2026 and Google has not announced it. Coverage says to ask a Google representative for access. Channel grouping differs between accounts, and there is no documented API field and no published scale for each step.

Can I run Conversion Lift on Performance Max myself now?

Google's setup page now lists Search and Performance Max as supported. You need at least 1,000 observed conversions, a study budget of at least $5,000 and a compatible conversion action. The same page still says Conversion Lift is not available to all accounts, so access may still run through a representative.

Sources

1. Search Engine Land, Google tests channel prioritization controls for Performance Max searchengineland.com
2. Search Engine Roundtable, Google Ads Channel Settings Prioritization Slider (Alpha Test) seroundtable.com
3. TechWyse, Google Tests PMax Channel Priority Controls techwyse.com
4. PPC News Feed, Performance Max Adds Channel Prioritization Controls ppcnewsfeed.com
5. MetricFixer, Google Tests Channel Prioritization Controls in Performance Max metricfixer.com
6. Spilno Agency, Performance Max channel controls alpha in Google Ads spilnoagency.com.ua
7. Brainlabs, Google Ads Is Testing PMax Channel Controls. Should Advertisers Use Them? brainlabsdigital.com
8. Veritas Digital, Google Performance Max Channel Controls 2026 veritas-digital.co.uk
9. Google Ads Help, About the channel performance report for Performance Max support.google.com
10. Google Ads Help, Evaluate Performance Max results support.google.com
11. Google Ads Help, Channel controls in Demand Gen campaigns support.google.com
12. Google Ads Help, Changes to target based bid strategies support.google.com
13. Google Ads Help, Frequently asked questions about changes to Target-based bid strategies support.google.com
14. Google Ads Help, Set up Conversion Lift based on users support.google.com
15. Google Ads Help, Comparing lift types support.google.com
16. Google Ads Help, About Performance Max experiments support.google.com
17. Google Ads Help, About conversion value rules support.google.com
18. Google, Channel performance and more reporting coming to Performance Max blog.google
19. Search Engine Land, Google opens Conversion Lift to Search and Performance Max searchengineland.com
20. Digiday, Google quietly gives ground on PMax controls digiday.com
21. PPC Land, Performance Max lets advertisers pick partner networks in alpha test ppc.land
22. TNT Growth, Performance Max Channels: 53-Account Spend Benchmarks tntgrowth.com
23. Optmyzr, Google Ads for BFCM 2026: What 4,200+ Accounts Reveal optmyzr.com
24. Optmyzr, Is Performance Max Cannibalizing Your Search Campaigns? optmyzr.com

NEXT STEP

See where your brand stands.

Your Digital Authority Score, who AI recommends for your target questions, and the placements that close each gap. Your first report is free.

Get your free report →