- Google now assigns customer types to conversion-based lists automatically, and you cannot leave an eligible list unclassified.
- That label feeds Smart Bidding, which pays more for clicks it believes lead to first-time acquisition.
- It misfires on lead generation, where a conversion is a form fill, not a customer. Audit every conversion-based list in Audience Manager.
What changed on 18 August
Google Ads now assigns customer types to audience lists built from conversion data, automatically. Advertisers can no longer leave an eligible list unclassified. Nothing needs switching on. It simply began happening to conversion-based lists across accounts.
The classification decides whether Google considers the people on a list new customers, returning customers, or something else. That is not a labelling exercise for tidiness. It feeds Smart Bidding.
Why a label changes your bids
Smart Bidding treats a new customer as worth more than a returning one, so it will pay more for a click it believes leads to first-time acquisition. Returning customers are cheaper to convert, so bids settle lower.
That logic is sound when the label is right. It is exactly as wrong as the label when it is not — and the failure is silent, because nothing in the interface tells you Google's judgement disagrees with yours.
The problem for lead generation
Here is where it stops being tidy. The classification is applied to lists built from conversions. In lead generation, a conversion is a form fill.
A form fill is not a customer. It is someone who might become one in six weeks after two calls and a proposal. When Google sorts that list into new and returning customers, it is describing a thing the business does not measure that way, and then adjusting bids on the strength of it.
This lands hardest on exactly the accounts least equipped to notice: B2B, professional services, anything with a sales cycle longer than a checkout. If your conversion action is "enquiry", every customer-type judgement downstream inherits a category error.
It also breaks in ordinary ways for ecommerce. The same household ordering under two email addresses reads as two new customers. A business buying under a personal address then a company one reads as new twice. Long repurchase cycles look like acquisition. None of that is new — what is new is that it now moves money.
What to audit this week
Open Audience Manager and go through every list built from conversion data. Three questions per list:
- 1. What type has Google assigned? Read it before assuming it is sensible. This is the number that is now driving bid aggressiveness.
- 2. Does the underlying conversion represent a purchase, or an enquiry? If it is an enquiry, the customer-type framing does not apply to your business and you should treat the resulting bid adjustments with suspicion.
- 3. Does Google's definition of a new customer match yours? Most businesses have a working definition — first paid order, first signed contract, first appointment attended. Google's is derived from whatever you told it a conversion was.
Start with the lists attached to your highest-spend campaigns. A misclassification on a list that governs 5% of budget is a curiosity. On the list governing 60% of it, it is a monthly variance you will spend a week explaining.
What to watch in the data
The signature to look for is a shift in cost per acquisition on campaigns using Customer Match audiences, starting around 18 August, without a corresponding change in volume or competition. If bids rose because Google decided a list was full of new customers, CPA rises while conversion volume stays flat.
As with any platform change, one campaign moving proves nothing — late August has its own seasonality. The signal is several Customer Match campaigns moving together while campaigns without those audiences hold steady.
Comparing before and after a date across every campaign and account is tedious by hand and trivial to ask for. 1ClickReport reads Google Ads alongside Meta, GA4 and Stripe, so “what changed after 18 August across all my accounts” is one question rather than an afternoon of exports. You can try it free.
The bigger picture
This is the same pattern as the bidding change a day earlier: a setting that used to be yours becomes an inference that is Google's. Individually each is defensible and often an improvement. Cumulatively, the number of levers an advertiser holds keeps shrinking, and the ones that remain matter more.
The practical response is not to fight it. It is to make sure the inputs Google infers from are correct, because inference on bad inputs is now the thing spending your budget.
Frequently asked questions
What changed in Google Ads customer lists on 18 August 2026?
Google began automatically assigning customer types to lists built from conversion data. Eligible conversion-based lists can no longer be left unclassified, and the assigned type feeds Smart Bidding.
Why does the customer type label affect my bids?
Smart Bidding weighs a new customer as more valuable than a returning one, so it pays more for clicks it believes lead to first-time acquisition. An incorrect label produces an incorrect bid adjustment.
Is this a problem for lead generation accounts?
It can be. The classification applies to lists built from conversions, but in lead generation a conversion is a form fill, not a customer. The customer-type framing describes something the business does not measure that way.
What should I audit?
Every list in Audience Manager built from conversion data: which type Google assigned, whether the underlying conversion is a purchase or an enquiry, and whether Google's definition of a new customer matches yours.
Sources: Google Ads Help — about audience customer types; Search Engine Land.
Related reading: the 17 August target bidding change, the Customer Match API migration, and audit a Google Ads account in five prompts.