Short answer: Run Search first if your account has no conversion history, a small product range, or a budget that produces only a handful of conversions a week. Search shows you the actual queries, so you learn what people type before you hand the budget to automation. Add Performance Max later, and keep brand traffic out of it.
What is the real difference between Performance Max and Search?
A Search campaign buys queries. You pick the keywords, you read the search terms report, you write the text ads, and the ad shows on Google Search.
Performance Max (PMax) buys everything at once, Search, Shopping, Display, YouTube, Gmail, Discover and Maps inventory, from one campaign, driven by your product feed, your asset groups and your audience signals. There are no keywords to bid on.
So the question is not which one performs better. It is whether you still need to learn what people type, or whether you already know and now want Google to spread the budget for you.
Which one should you run first on a small budget?
In almost every account without reliable conversion history, start with Search. Automation learns from conversions, and Search is the cheapest way to produce the first ones while leaving evidence you can actually read.
| Your situation | Start with | Why |
|---|---|---|
| Conversion tracking not verified | Neither | Both bidding systems learn from your conversion data. Bad data trains the wrong behaviour. |
| New account, no conversion history | Search | You need search terms before you can judge anything. PMax will not give you that detail. |
| Small catalogue, few conversions per week | Search, plus Standard Shopping if you sell products | The budget is too thin to feed a campaign spreading across six networks. |
| B2B leads, long sales cycle, low volume | Search | Intent sits in the query. Broad inventory mostly adds cheap, unqualified form fills. |
| Months of conversion data and a large feed | Performance Max | Enough signal to learn from, and enough products for automation to find demand you would never keyword. |
| Strong existing brand demand | Brand Search, kept in its own campaign | Keeps brand conversions out of the numbers you use to judge acquisition. |
What breaks when both run on the same products?
The two campaigns are not walled off from each other. When a query could be served by both, one rule matters most: if the query matches a keyword in your Search campaign exactly, the Search ad is the one eligible to serve. Otherwise Ad Rank decides, and PMax often wins. The current rules are documented at support.google.com/google-ads.
The damage is rarely burned money. It is that your reporting stops telling you where the growth came from.
| Symptom | Cause | Fix |
|---|---|---|
| PMax ROAS (return on ad spend) looks excellent, total revenue is flat | PMax is absorbing brand and existing demand | Apply brand exclusions and account-level negative keywords, run brand as its own Search campaign |
| Search impressions fall after the PMax launch | Overlap, PMax wins whatever is not an exact-match keyword | Tighten Search to exact match on proven terms and judge results at account level |
| Standard Shopping spend collapses | PMax takes priority for the same products | Decide deliberately which campaign owns which products instead of letting the auction decide |
| You cannot explain a bad month | No search term history, because Search was never run | Keep a small Search campaign alive for diagnostics even when PMax carries the spend |
How to launch them in the right order, step by step
- Fix measurement first. One primary conversion action per goal, no duplicates, imported cleanly into Google Ads. In the EU, confirm consent mode is live and know what share of your traffic consents, otherwise modelled conversions hide the problem instead of solving it.
- Calculate your own targets. Break-even ROAS = 1 ÷ gross margin. Say your margin is 40% on revenue excluding VAT: 1 ÷ 0.4 = 2.5, so 250% ROAS is break-even, not success. For leads: target cost per lead = profit per closed deal × your lead-to-customer rate × the share of that profit you are willing to spend.
- Launch Search in two parts. Brand in one campaign, generic in another. Start the generic one on exact and phrase match with terms you can defend, not broad.
- Mine search terms weekly. Add negatives, move budget to themes that convert. That list becomes your audience signals and asset group input later.
- Add Performance Max only once the account converts predictably. Clean feed, brand exclusions on, account-level negatives applied, asset groups split by category or margin.
- Compare at account level. Hold the budget steady and compare total revenue or leads over equal windows. MER (marketing efficiency ratio, total revenue ÷ total ad spend) tells you more here than campaign ROAS.
What I see in the accounts I run
Across 500+ campaigns and more than €5 million in managed ad spend, one pattern repeats more than any other. I take over an account where PMax shows the best ROAS in the interface, and the owner is convinced it is the strongest campaign they have.
Then I add brand exclusions and move brand into its own Search campaign. PMax ROAS drops. Total account revenue usually does not move. That gap is the campaign reporting demand it did not create.
The second pattern is diagnostic poverty. Accounts that launched with PMax and nothing else have no search term history at all, so when results slide there is nothing to look at except asset reporting and a budget slider. Accounts that ran Search first have a list of converting terms, and that list is exactly what makes a later PMax launch good: better signals, better asset groups, better negatives.
The third is size. In Slovakia, Czechia or Austria the addressable audience is a fraction of a US account. On a budget of, say, €1,500 a month, a campaign that spreads across six networks spends a large part of it learning. Search concentrates the same money on people already looking. Most of the accounts I audit as a PPC specialist have this backwards.
When not to run Performance Max at all
- Tracking is broken or largely unconsented. You would be training automation on noise.
- The website does not convert. If high-intent Search traffic does not buy, more channels will not fix it. The problem is price, product, delivery or the page.
- The feed is thin, wrong or half out of stock. PMax is a feed campaign first.
- You run both on the same products and compare them against each other. They share auctions, so that is not a test.
- You switch both on in the same week. Then no result can be attributed to either.
Start here: verify that conversions are tracked correctly, then launch one brand Search campaign and one generic Search campaign with terms you can defend. Give it a month of search term mining before Performance Max touches the same products.
If you want that structure built and run properly, see Google Ads management or the full list of services. Email me and I’ll reply within 24 hours, write to me here.
FAQ
Can I run Performance Max and Search at the same time?
Yes, and most mature accounts do. Keep brand in a separate Search campaign, apply brand exclusions and account-level negative keywords to PMax, and judge the result at account level rather than comparing the two campaigns.
Does Performance Max steal traffic from my Search campaigns?
It competes for the same queries. If the query is an exact match to a keyword in your Search campaign, the Search ad is eligible to serve; otherwise Ad Rank decides and PMax often wins.
How much budget do I need before Performance Max makes sense?
There is no universal number. The useful test is conversion volume, not spend: if your account produces only a few conversions a week, automation across six networks has too little signal to learn from.
Should a B2B company with long sales cycles use Performance Max?
Usually not first. Start with Search, feed qualified-lead or closed-deal values back into the account, and only then consider PMax. Otherwise it optimises towards whichever form fills are cheapest, not best.
Why does my Performance Max ROAS look so good?
Often because it is capturing brand searches and returning customers. Add brand exclusions, split brand into its own Search campaign, and check whether total account revenue actually changed.
Do I still need Standard Shopping if I run Performance Max?
Not necessarily, but decide it consciously. PMax takes priority over Standard Shopping for the same products, so leaving both running usually means the Shopping campaign quietly stops spending.