Short answer: I leave Performance Max’s new customer acquisition setting off for ecommerce accounts. It only works if Google can tell your new customers from your existing ones, and for most stores that matching is patchy. When I tested it on an ecommerce account for a month, the effect on performance ranged from neutral to negative, so I switched it off. To push spend toward new customers, I exclude brand terms from Performance Max and let a separate Search campaign handle brand.
What the new customer acquisition setting does
New customer acquisition is a customer lifecycle goal in Google Ads. It changes how much Smart Bidding will pay for a conversion depending on whether Google thinks the buyer is new or existing. There are three modes:
| Mode | What it does | Bid strategies | Campaign types |
|---|---|---|---|
| New customer value | Bids higher for new customers than for existing ones | Maximise conversion value or Target ROAS | Search, Performance Max, Shopping, Demand Gen |
| High-value new customer | Separate bid priorities for high-value new, regular new and existing customers | Maximise conversion value or Target ROAS | Performance Max, Search |
| New customer only | Bids only for people Google classes as new | Value-based, or Maximise conversions / Target CPA | Search, Performance Max, Shopping, Demand Gen |
Google recommends “New customer value” for all advertisers with purchase goals. It recommends “New customer only” only for advertisers with strict acquisition budgets, and suggests a separate campaign for existing customers if you use it. (Google Ads Help)
The assumption it rests on
Every mode depends on one thing: Google correctly sorting each buyer into “new” or “existing”. Google asks you to define existing customers with first-party data, a customer list and a website tag.
That is where it breaks down for most stores:
- Customer lists go stale. A list uploaded once and not refreshed misses everyone who bought since.
- Matching is partial. Google can only match customers whose details it can link to a Google account. Guest checkouts, different email addresses and work-versus-personal accounts all slip through.
- Consent and cookies cut the tag’s view. Visitors who decline tracking, clear cookies or switch devices can look new when they are not.
When the classification is wrong, the bidding is wrong in both directions. The campaign pays a premium for some existing customers it thinks are new, and holds back on some genuinely new customers it thinks it has already seen.
What happened when I tested it
I ran the setting on an ecommerce account for one month. The effect on account performance was neutral at best and negative at worst. It didn’t lift the result it was meant to lift, so I switched it off and have kept it off since.
One month is not a long test, and a different account with cleaner customer data might see something else. But the result matched the problem above: the setting is only as good as Google’s view of who your customers are, and that view is rarely complete enough to bid on.
What I do instead: exclude brand from Performance Max
The most reliable signal that someone already knows your store is that they searched for your brand. So instead of asking Google to guess who is new, I take brand traffic out of Performance Max:
- Add your brand terms as a brand exclusion in the Performance Max campaign settings. Include common misspellings.
- Run brand in its own Search campaign. Branded searches are cheap and convert well. Keeping them separate stops them inflating Performance Max’s reported ROAS.
- Let Performance Max spend on non-brand demand. With brand removed, its budget goes to people searching for products and categories, which is where most new customers come from.
- Report brand and non-brand separately. Judge Performance Max on non-brand revenue, not blended ROAS. Brand vs non-brand ROAS covers how to split the numbers.
This isn’t a perfect proxy either. Some new customers search your brand after seeing an ad or a social post, and brand exclusion moves them to the Search brand campaign. But the rule is simple, it doesn’t depend on customer matching, and it changes where the money goes from day one.
Brand exclusion works alongside the rest of the structure: campaigns split by margin and budgets set from break-even. The Performance Max for ecommerce structure guide and the Performance Max bidding guide cover those layers.
When the setting might still be worth testing
I keep it off, but there are accounts where a test is reasonable:
- Stores with strong first-party data, where most buyers log in and the customer list is refreshed automatically.
- Accounts with a strict acquisition budget, which is the case Google itself recommends “New customer only” for.
- Subscription or high-repeat brands, where a new customer is worth far more than a repeat order and you can put a real number on the difference.
How to test it properly
- Fix the data first. Upload a current customer list, set it to refresh automatically if your platform allows, and confirm the Google tag fires on purchase.
- Change one thing. Don’t restructure campaigns or switch bid strategies during the test.
- Run at least four weeks at steady budget, so the bidding has time to adjust.
- Judge it on your own numbers, not Google’s labels. Compare new customer orders in Shopify or your store platform, non-brand revenue and overall marketing efficiency against the month before. Google’s “new customer” column uses the same classification you are testing.
- Switch it off if the result is flat or worse. A setting that doesn’t improve the outcome still adds a layer you have to explain in every report.
Running Performance Max for new customer growth without overpaying for existing ones is part of ecommerce Google Ads management at Adgenix.
Frequently asked questions
What is new customer acquisition in Google Ads? A customer lifecycle goal that tells Smart Bidding to bid more for new customers, or only for new customers. It is available in Search, Performance Max, Shopping and Demand Gen campaigns.
Does new customer acquisition work in Performance Max? It depends on how well Google can identify your existing customers. In my one-month test on an ecommerce account the effect was neutral to negative, so I keep it off.
What is the difference between “new customer value” and “new customer only”? New customer value bids higher for new customers but still bids for existing ones. New customer only bids for new customers alone, and Google recommends it only for strict acquisition budgets.
How does Google know who is a new customer? From the first-party data you provide, mainly customer lists and the website tag. If those are incomplete or out of date, the classification is too.
Is excluding brand terms the same as new customer acquisition? No, but it targets a similar goal more predictably. People searching your brand mostly know you already. Excluding brand from Performance Max moves its spend to non-brand searches, where new customers are more likely to come from.
Sources
