Performance Max Campaigns: When They Earn Budget and When They Burn It

Performance Max is not the right answer for every Google Ads account. It earns budget when an account has retail or ecommerce inventory, enough conversion volume to train the algorithm, and a Search and Shopping foundation that PMax can sit on top of. It burns budget on small accounts, lead-gen with weak signals (a single conversion event with no quality grade tied to revenue stages), and B2B account-based motions. The decision is whether PMax fits the account, not whether PMax works.

PMax is easier to manage than a structured Search and Shopping account, and that operational reality drives a lot of agency recommendations. PMax is one campaign, one goal, one budget. Search and Shopping take feed work, negative keyword sculpting, asset group structure, and weekly bid tuning. PMax can be turned on in an afternoon. Agencies that scale account volume per media buyer save hours by switching client accounts to PMax.

This does not necessarily make PMax the wrong answer for the account. The more crucial question is whether the agency picked PMax based on your data or by default for every client. Choosing PMax for operational ease rather than account fit is one of the common ways performance marketing programs go sideways.

Performance Max is a single Google Ads campaign type that runs across Search, Shopping, Display, YouTube, Gmail, Discover, and Maps inventory inside one campaign goal [1]. The advertiser supplies assets, audience signals, and a conversion goal. Google controls targeting, creative mix, and placement. PMax replaced Smart Shopping and Local campaigns in late 2022.

What Google advertises: one campaign, all inventory, algorithmic optimization. What Google does not lead with: no search-term visibility by default, no placement-level bid control, no way to exclude Display or YouTube from the mix, and limited transparency into where spend lands.

Google has added more granular controls over time, such as a search-term report rolled out in 2024 that covers Search and Shopping inventory inside PMax. Display, YouTube, Gmail, and Discover placements remain dark, so visibility is partial.

It is also worth saying plainly why Google promotes PMax so hard. PMax expands the surfaces Google can monetize inside one campaign goal, increases automation revenue per advertiser, and reduces the need for human campaign management. None of that is conspiratorial. PMax is a real product with a real business model. Knowing that incentive helps you read agency and Google rep advice with clearer eyes.

PMax earns budget when three conditions hold: the account has enough conversion volume to train the algorithm, the business has retail or ecommerce inventory Google can match to intent, and Search is already maxed out on qualified keywords. In those scenarios, PMax is an efficient assist layer that complements Search and Shopping rather than replacing them.

When PMax usually works well:

• High-volume ecommerce with at least 20 to 30 conversions per month per campaign.
• Accounts where Search already captures every qualified keyword and the next dollar of growth has to go somewhere.
• B2B only when sales cycles run 60 days or longer, multi-touch attribution is in place, and value-based bidding feeds CRM data back to Google.
• Best treated as an assist layer that complements Search and Shopping rather than replacing them.

Google’s own positioning makes the role clear: PMax is designed to use one goal across multiple Google surfaces, not to replace every campaign type in the account. That argues for PMax as a layer on top of a working Search and Shopping foundation, not as a replacement for those campaigns. The same logic shows up in ROAS math that ties to the P&L on the ecommerce side.

PMax breaks down on small budgets, weak conversion signals, B2B account-based motions, and anything that depends on precise keyword targeting. One audit of a B2B SaaS account running under $20,000 a month in Google spend showed the pattern clearly. PMax pulled in branded traffic and existing-customer cookies, inflated platform ROAS, and hid what was happening in non-brand acquisition.

When PMax does usually not work:

• Budgets under $1,000 a month: not enough volume for the algorithm to learn. Display and YouTube eat the budget without converting.
• Lead-gen with a single conversion action and no quality signal: Google finds the cheapest form fills, which tend to be the worst leads.
• B2B account-based motions with a few hundred named accounts: the algorithm does not have enough room to learn or scale.
• Regulated industries where auto-generated assets can trigger policy violations.

The cannibalization pattern is the most consequential failure mode and the easiest to miss from a platform dashboard. A concrete example makes it real. An ecommerce account runs an exact-match Search campaign on “leather messenger bag.” The same account runs a PMax campaign with the full product feed. A user types the query. Both campaigns are eligible. Google’s auction picks one, and outside narrow exact-match exceptions, PMax wins. PMax serves the ad, the user clicks, the conversion is logged to PMax. The advertiser opens the dashboard, sees PMax revenue rising and Search revenue falling, and concludes PMax is working. In reality, PMax is harvesting demand Search already had, often at a higher cost per click and a lower conversion rate.

The risk is structural. PMax and Search can both be eligible around the same commercial intent, while campaign reporting attributes the click to whichever campaign won the auction. That means a dashboard can show PMax revenue rising and Search revenue falling without proving incremental demand. That is what an account audit surfaces that a platform dashboard hides.

Most PMax decisions get made on native platform ROAS, which understates the truth by at least half. Across every account audited in Opascope’s portfolio, native attribution underreports incremental revenue by 50% or more. PMax is hit hardest because it reports across Display, YouTube, Gmail, and Maps, where most engagement is view-through or assisted rather than direct-click. Click-based attribution misses that contribution entirely, undercounting revenue PMax drove. At the same time, the click-based model over-credits last-click channels like Search, Direct, and branded organic. PMax looks better or worse than it is depending on which side of the model is wrong, and the advertiser cannot tell from platform data alone.

Quick takeaways:

• Native conversion counts reflect what Google can model, not what the business sees in revenue.
• PMax’s cross-channel surface inflates view-through credit on Display and YouTube, and native attribution does not tie those views to incremental revenue.
• Across the portfolio, platform ROAS often reads above 1,000% while the P&L view comes in below 100%.
• Fix measurement first. A 13-week spend-to-revenue lag, Marketing Efficiency Ratio, and incremental CAC beat any attribution model.

The short version of how we approach paid media measurement: read native ROAS as a directional signal, not a verdict.

For most accounts, the right split is Search and Shopping first, PMax second. For ecommerce at scale, a 60/40 or 50/50 split between Shopping plus Search and PMax tends to hold. For B2B lead generation, PMax rarely earns more than 20 to 30% of the Google budget, and only with value-based bidding pulling CRM lead quality into the signal.

Quick takeaways:

• Budget logic by account size. Under $10,000 a month in Google spend, run Search and Shopping only. From $10,000 to $50,000, test PMax at 15 to 25%. Above $50,000, scale PMax with guardrails.
• Campaign role clarity. Search owns high-intent and brand. Shopping owns product-intent. PMax owns broader discovery and warm retargeting.
• Guardrails on PMax. Cap spend, apply brand exclusions, feed first-party audience signals, and turn on value-based bidding where the data supports it.
• Multi-channel structures consistently beat single-campaign accounts when each campaign type has a defined role rather than competing for the same query.

If PMax is already running and leadership is not killing it, the highest-impact fixes are value-based bidding with CRM data, brand-term exclusions, audience signals built from first-party lists, and asset group structure that mirrors the actual product taxonomy. Fixing attribution sits underneath all of it.

Quick takeaways:

• Pipe CRM lead quality into PMax with offline conversion imports. An offline conversion is any event that happens outside Google Ads tracking, like a sales-qualified lead in the CRM, a closed-won opportunity, or a contract signed, pushed back into Google Ads via Enhanced Conversions for Leads, the Google Ads API, or a CRM connector. The advertiser sends Google a conversion value tied to the original click ID so the bidding model trains on what became revenue, not just what filled out a form. This is the single highest-impact PMax fix. A raw form-fill signal trains the algorithm to find cheap form fills. A revenue-stage signal trains it to find buyers [2].
• Exclude brand terms using the brand-list feature so PMax stops taking credit for demand the account already has.
• Build audience signals from first-party customer lists, not lookalikes alone. First-party data beats default targeting on every account we have audited.
• Structure asset groups by product line, persona, or funnel stage. Generic asset groups are why creative quality scores stay low and Display eats budget.

This is the work behind one cloud platform that cut cost per acquisition by 93% and grew signups 660% over four years after we rebuilt attribution and aligned PMax to down-funnel signals. On the same engagement, paid went from producing almost no paying customers to 30 to 40% of new customers. Cost per lead rose 40 to 60%, and cost per paying customer dropped.

If your agency moved your Google Ads budget into PMax, the right questions are about structure and signals, not reporting. Ask what share of budget sits in PMax, whether brand terms are excluded, whether value-based bidding is live, and how platform ROAS compares to a blended view.

• What percent of Google budget sits in PMax, Search, and Shopping today, and why?
• Are brand terms excluded from PMax? Show me the exclusion list.
• Is value-based bidding live? What conversion values are assigned to lead stages or order revenue?
• What does platform ROAS say versus our blended P&L view, and where do they diverge?
• What audience signals feed PMax, and where do they come from?
• Why PMax instead of Search or Shopping for this specific objective?

If the answers are vague, that is the answer. For a second read on your account, start here.

What are Performance Max campaigns?

Performance Max is a Google Ads campaign type that runs across Search, Shopping, Display, YouTube, Gmail, Discover, and Maps inventory in one campaign. Google controls targeting, creative mix, and placement. The advertiser supplies assets, audience signals, and a conversion goal. PMax replaced Smart Shopping and Local campaigns in late 2022.

Should I use Performance Max or Search campaigns?

Start with Search and Shopping for high-intent queries. Add PMax once Search inventory is fully captured and the account has at least 20 to 30 conversions per month per campaign. For B2B lead generation or budgets under $10,000 a month, Search alone usually wins. PMax works best as an assist layer on top of a structured Search and Shopping foundation.

When should I avoid Performance Max?

Avoid PMax on budgets under $1,000 a month, on B2B account-based motions targeting a small named list, on lead-gen campaigns with no lead-quality signal, and in regulated industries where auto-generated assets can trigger policy reviews. In each case, the algorithm either lacks volume to learn or learns the wrong thing because the conversion signal does not match revenue.

Does Performance Max cannibalize Search campaigns?

Yes, Performance Max can cannibalize Search campaigns when both campaign types are eligible around the same commercial intent. The symptom is simple: PMax reported revenue rises, Search reported revenue falls, and total incremental revenue does not move enough to justify the shift. The fix is to separate campaign roles, apply brand exclusions, and judge the change against blended revenue rather than campaign-level ROAS alone.

What is the minimum budget for Performance Max?

The practical floor is around $1,000 a month for ecommerce with a Shopping feed. Below that, Display and YouTube consume budget without generating qualified traffic. For PMax to perform well, most accounts need 20 to 30 conversions per month per campaign.

Is Performance Max good for B2B lead generation?

PMax works for B2B only when sales cycles run 60 days or longer, value-based bidding is live with CRM lead quality feeding back into Google, and the addressable market is large enough for the algorithm to learn. For tightly targeted account-based motions with a few hundred named accounts, PMax is usually the wrong tool. The algorithm needs scale, and ABM lists do not give it scale.

Why are my Performance Max leads low quality?

PMax bids toward whatever conversion signal you feed it. If the only conversion is a form fill with no quality score, the algorithm finds the cheapest form fills, which tend to be the worst leads. The fix is to import offline conversion data from the CRM and assign value to lead stages like MQL, SQL, Opportunity, and Closed-Won. An offline conversion is an event outside Google Ads, like a CRM stage change or a closed-won deal, pushed back into Google Ads through Enhanced Conversions for Leads or the Google Ads API. That lets PMax train on what becomes revenue.

How do I exclude brand terms from Performance Max?

Google added a brand-list exclusion feature to PMax in 2024. Build a brand list inside account settings, apply it to the PMax campaign, and verify that branded search terms no longer appear in the PMax search-term report. Excluding brand terms prevents PMax from harvesting demand the account already has.

Can I see search terms in Performance Max?

Google added a search-term report to PMax in 2024. It covers Search and Shopping inventory inside PMax but not Display, YouTube, Gmail, or Discover placements, so visibility remains partial. For the dark surfaces, asset-group reporting and placement reports are the closest available signal.

What is the right budget split between Performance Max and Search?

It depends on the account. For ecommerce at $50,000 a month or more in Google spend, a 40 to 60% allocation to PMax with the rest in Search and Shopping is common. For B2B lead generation, PMax rarely earns more than 20 to 30% of the Google budget. Below $10,000 a month, PMax is usually not the right starting point.

Why did my agency move everything to Performance Max?

PMax takes less time to manage than Search and Shopping, so for agencies scaling account volume per media buyer it is a convenient default. That does not make it the right answer for the account. Ask which campaign type best fits the objective, and ask to see the data that informed the recommendation.

How do I know if Performance Max is actually working?

Platform ROAS is not enough. Compare PMax-reported conversions against a blended view: Marketing Efficiency Ratio, a 13-week spend-to-revenue lag, and incremental CAC. Across every account we audit, native attribution underreports by 50% or more. If PMax shows strong ROAS on platform but the P&L does not reflect it, the platform number is wrong.

1. Google Ads Help. About Performance Max campaigns. https://support.google.com/google-ads/answer/10724817
2. Google Ads Help. Performance Max best practices for lead generation. https://support.google.com/google-ads/answer/13775965