IBRAVRA Media Network
Marketing • 11 min read

Google Performance Max vs Search Campaigns: ROAS, Control & the Break-Even Line

Performance Max and Search are opposite control models, not two versions of the same campaign. Search lets you choose the query, the ad, and the landing page; Performance Max lets Smart Bidding choose all three from your assets and goals. This playbook runs both through the break-even ROAS formula — one divided by gross margin — with a $5,000 worked example, and gives the setup sequence that actually works.

By Enow A. Jovial • Published 2026-11-17
Performance Max and Search are not two versions of the same campaign — they are opposite control models. Search lets you choose the query, the ad, and the landing page; Performance Max lets Google's Smart Bidding choose all three from the assets and goals you supply. At the same budget, the winner is decided by one number: your break-even ROAS, which is one divided by your gross margin. A 40% margin business breaks even at 2.5x ROAS; below that, neither campaign type saves you — above it, the campaign that holds its target cheapest wins.

What Performance Max is — and what it is not

Performance Max is a goal-based campaign type that lets one campaign access all of Google Ads inventory — YouTube, Display, Search, Discover, Gmail, and Maps — from a single campaign. That sentence, from Google's own help page, is the whole architecture: one budget, one goal, every surface. You supply the conversion goals and an optional ROAS or CPA target, you supply text, images, and video assets, and Smart Bidding sets a bid for every query at auction time based on the probability that the query converts. Google's AI combines your assets into ads, generates additional headlines and descriptions from your landing page content, and funnels your budget toward the highest-performing inventory and placements.

What it is not is a keyword campaign with extra reach. There are no keywords to bid on, no match types to tune, and no per-query ad copy to write. The control surface is the asset group, the audience signal, the conversion goal, the daily budget, and the brand-safety settings — brand exclusions, placement exclusions, and content-suitability filters. If your advertising skill is writing the exact ad for the exact query, Performance Max removes the place where that skill applies. If your skill is feeding a system good creative and a clean conversion signal, it multiplies it.

Google positions Performance Max as a complement to keyword-based Search campaigns, not a replacement for them, and that framing is the honest one. The two campaign types answer different questions. Search answers "who is already looking for exactly this?" Performance Max answers "who, anywhere on Google's inventory, is most likely to convert at the price I set?" A store that runs only one of them is answering only one of those questions.

What a Search campaign controls that Performance Max does not

A Search campaign gives you four levers Performance Max does not offer. First, the query: you choose the keywords, the match types, and the negative keywords, so you can refuse to pay for a query entirely. Second, the ad: you write the headline and description that a specific query sees, so a query for "emergency plumber" and a query for "plumber price list" never share copy. Third, the landing page: each ad group points at the page that closes that query. Fourth, the bid: you can set manual bids per keyword or a portfolio bid strategy across campaigns, and you can see the search terms report and add the exact terms that converted as keywords.

Performance Max replaces those four levers with two: the asset group and the goal. An asset group is a bundle of text, images, and video that Google combines into ads across every channel, with optional audience signals that guide the system toward high-value intent. Final URL expansion, if you turn it on, lets Google match queries to landing pages on your site and generate dynamic assets for them — which means Google can send a click to a page you never chose for that query. You remain responsible for reviewing the landing page content and every dynamically generated asset, and that responsibility is the single most common source of wasted Performance Max spend: a system allowed to expand URLs will expand them to the pages that convert cheapest, not the pages that sell your best product.

ControlSearch campaignPerformance MaxAuthority
Query targetingKeywords, match types, negativesNone — goal + audience signals onlyGoogle Ads Help, About Performance Max campaigns
Ad copy per queryWritten per ad groupAI-combined from asset groupGoogle Ads Help, About Performance Max campaigns
Landing page choicePer ad group, chosen by youFinal URL expansion may chooseGoogle Ads Help, About Performance Max campaigns
Bid per queryManual or portfolio Smart BiddingAuction-time Smart Bidding onlyGoogle Ads Help, About Smart Bidding
InventorySearch + Search partnersYouTube, Display, Search, Discover, Gmail, MapsGoogle Ads Help, About Performance Max campaigns
Brand safetyNegative keywordsBrand exclusions, placement exclusionsGoogle Ads Help, About Performance Max campaigns
ReportingSearch terms, per keywordAsset and audience aggregatesGoogle Ads Help, About Performance Max campaigns

The ROAS math: one formula decides the winner

Return on ad spend is revenue divided by spend. Break-even ROAS is the ROAS at which the ad spend consumes exactly the gross profit the sales produce, and it depends on nothing but margin:

Break-Even ROAS=1Gross Margin\text{Break-Even ROAS} = \frac{1}{\text{Gross Margin}}

A business with a 40% gross margin breaks even at 2.5x. A 25% margin business breaks even at 4.0x. A 60% margin software business breaks even at 1.67x. Before you compare campaign types, compute this number — it is the only target that matters, and every Target ROAS you set should be above it by the profit you intend to keep. A Target ROAS of 3.0x at 40% margin leaves 25% of revenue as gross profit after ad spend; a Target ROAS of 2.5x at the same margin leaves nothing.

Run the two campaign types through the same arithmetic with your own account's numbers. The figures below are illustrative inputs, clearly labelled as such — replace every one with your account's actuals before you decide anything. Assume a $5,000 monthly budget, an average order value of $120, and a 40% gross margin, so break-even ROAS is 2.5x.

For a Search campaign at an average cost per click of $2.50 and a 3.0% conversion rate, the cost per acquisition is the click price divided by the conversion rate:

Search CPA=$2.500.030=$83.33\text{Search CPA} = \frac{\$2.50}{0.030} = \$83.33
Search ROAS=$120$83.33=1.44x\text{Search ROAS} = \frac{\$120}{\$83.33} = 1.44x

For a Performance Max campaign at a blended cost per click of $1.80 and a 2.5% conversion rate, the same arithmetic gives:

PMax CPA=$1.800.025=$72.00\text{PMax CPA} = \frac{\$1.80}{0.025} = \$72.00
PMax ROAS=$120$72.00=1.67x\text{PMax ROAS} = \frac{\$120}{\$72.00} = 1.67x

Performance Max wins the comparison and still loses money: both campaigns sit below the 2.5x break-even line, so at a $120 average order value this product cannot carry paid search at either click price. The fix is not a different campaign type — it is a higher order value, a better conversion rate, or a cheaper click. Raise the average order value to $200 with the same click prices and conversion rates, and the picture reverses: Search ROAS becomes 2.40x and Performance Max becomes 2.78x, so Performance Max clears break-even and Search does not. The campaign type never changes the break-even line; it only changes where you land relative to it.

Two more numbers complete the comparison. Monthly conversions at the $5,000 budget are spend divided by CPA: about 60 for Search and about 69 for Performance Max in the illustration above. And the learning cost is real: Smart Bidding sets bids from conversion history, so a campaign with almost no conversions has almost nothing to learn from. A new account that launches Performance Max first is asking the system to optimize on a handful of data points; a new account that runs Search first, collects conversions against known queries, and then adds Performance Max is handing the system a trained signal. Sequence matters more than settings.

When Performance Max wins, and when Search wins

Performance Max wins when the product is visual, the audience is broad, and the conversion signal is strong: ecommerce with a deep product feed, a proven offer, and hundreds of monthly conversions. It wins when the customer journey crosses channels — a shopper who sees a YouTube ad, reads a Discover card, and searches the brand name a day later is one conversion to Performance Max and three separate campaigns to you. It wins when your catalogue is too large to write keyword-level ads for every product, because the asset group scales where ad groups do not.

Search wins when intent is narrow and expensive: a high-ticket B2B service where one click costs $40 and the wrong query burns a day's budget, a regulated product where the ad copy must say exactly one thing, a brand-defence play where you must own your own name. Search wins when the query itself is the qualification — "emergency", "near me", "same day" — because only Search lets you bid on that word and refuse the rest. And Search wins whenever you need to know why a click converted, because the search terms report names the query and Performance Max reports aggregates.

The common failure is running Performance Max as a Search replacement on a thin signal: a new store, a $20 daily budget, five conversions a month, and a Target ROAS copied from a blog post. The system spends the budget, the ROAS lands below break-even, and the owner concludes Performance Max does not work. What did not work was the setup — too little conversion data, too low a budget for the target, and a target below the break-even line. The campaign type is rarely the defect; the arithmetic usually is.

Final URL expansion is the quietest way Performance Max wastes money. With it on, Google can match a query to any page on your site and generate the ad for it — including clearance pages, blog posts, and pages whose prices changed months ago. Audit the landing pages Performance Max actually served before you raise the budget, and turn expansion off for any catalogue where the margin varies sharply between products.
Never put a 60% margin product and a 20% margin product in one Performance Max campaign with one Target ROAS. The system will chase the cheap conversions on the low-margin product and report a healthy blended ROAS while the profit disappears. Split asset groups by margin band, set the target above each band's own break-even line, and judge each band on its own ROAS.

The setup sequence that actually works

Step 1 — Run Search first. Build the keyword list, write the ads, collect thirty days of conversions against known queries, and compute your real cost per click, conversion rate, and ROAS from the account — not from a benchmark article.

Step 2 — Add Performance Max second. Use the same conversion goal, the same daily budget as a starting point, and a Target ROAS set above your break-even line by the margin you intend to keep.

Step 3 — Feed it proven assets. Use the headlines that already won in Search, the images from the best-selling product pages, and a video if you have one.

Step 4 — Set brand exclusions. Keep Performance Max off your brand name where Search already owns it, and set placement exclusions for the inventory you will not accept.

Step 5 — Audit Final URL expansion. Check the landing pages actually served before raising the budget; turn expansion off where margins vary sharply.

Step 6 — Compare on one metric. ROAS against the same break-even line, measured over the same thirty days, with the same attribution. The campaign that clears break-even by the wider margin gets the next dollar. The campaign that does not clear it gets a fix — a better offer, a tighter asset group, a higher order value — not a bigger budget. A budget increase on a sub-break-even campaign is a decision to lose money faster, and the only thing the campaign type changes is how fast.

For the cross-platform view, the Google Ads vs Meta Ads ROAS playbook runs the same break-even arithmetic across platforms, and the LinkedIn Ads vs Google Ads lead-cost guide covers the B2B case where the conversion is a lead, not a sale. Run the numbers through the ROI calculator before you set any Target ROAS — the calculator turns your margin and order value into the break-even line in one step.

Founder & CEO Insight — Enow A. Jovial

"Every advertiser I have watched lose money on Performance Max made the same mistake: they set the target from a benchmark instead of from their own margin. Your break-even ROAS is one divided by your gross margin. If you cannot state your gross margin, you are not ready for Performance Max — you are ready for a spreadsheet."
— Enow A. Jovial, Founder & Chief Executive Officer

Performance Max vs Search: the launch checklist

  • [x] Compute break-even ROAS: One divided by gross margin; write it down before opening Google Ads.
  • [x] Run Search first: Collect thirty days of conversions against known queries and record real CPC, conversion rate, and ROAS.
  • [x] Set the Target ROAS above break-even: Target equals break-even plus the profit margin you intend to keep — never below break-even.
  • [x] Split asset groups by margin band: One Target ROAS per margin band; never average a 60% product with a 20% product.
  • [x] Set brand exclusions: Keep Performance Max off your brand name where Search already owns it.
  • [x] Audit Final URL expansion: Check the landing pages actually served before raising the budget; turn expansion off where margins vary sharply.
  • [x] Compare on one metric: ROAS against the same break-even line, same thirty days, same attribution — then move the next dollar to the winner.