Customer Match is the Google Ads feature that lets an advertiser build audiences from its own first-party customer data - email addresses, phone numbers, names with postal addresses, and mobile device IDs - which Google matches against signed-in users so the advertiser can target, exclude or expand from them across Search, Shopping, YouTube, Gmail and Display. It is Google's version of the matched audience, with its own formats, its own policy gate and its own minimum sizes.
The definition
You upload a customer list, or push one through the Google Ads API from a connected account, with the identifiers formatted and hashed the way Google specifies. Google compares the hashes to the accounts of signed-in users and creates an audience of the users it recognized; you see the size, never the individuals. The audience can then be targeted, excluded, or used as the basis for expansion, and it applies across the Google surfaces that support it - which, because it includes Search, makes it one of the few ways to put a first-party list in front of people at the moment they are searching for your category.
What it matches on
Four kinds of identifier, each hashed before upload or by Google at upload: email addresses, phone numbers in international format, first and last names combined with country and postal code, and mobile device IDs. Google's documentation describes the exact normalization - lowercase, trimmed, specific formats for phones and names - and a value that deviates hashes to something that matches nothing, which is why normalization before hashing is the step that most affects match rates. Rows carrying several identifiers give Google more than one chance to recognize a user, and business email addresses generally match less often than personal ones because fewer people sign in to Google with them - the main reason B2B match rates run lower here than on LinkedIn.
Policy and size requirements
How B2B teams use it well
Three uses carry the weight. Suppression: excluding customers, open opportunities, employees and competitors from Search campaigns, so paid clicks stop going to people who would have found you anyway or should never be paid for. Retargeting with intent: open opportunities and signal movers from a standing watch reached on Search and YouTube while the conversation is live. Expansion seeds: closed-won customers and best-fit accounts as the basis for Google's own similar-audience expansion, with the suppression list applied so the seed teaches the right shape. All three depend on the list underneath being verified, normalized and refreshed on a schedule, which is the pipeline in AI agents for matched and custom audiences; AstroFabric's audience agent produces the platform-ready Customer Match export from any persistent list, with suppressions applied and the identifiers normalized and hashed.
Frequently asked questions
What is Google Customer Match?
A Google Ads feature that builds audiences from your own first-party customer data - hashed emails, phones, names with addresses, device IDs - matched against signed-in Google users, for targeting, exclusion or expansion across Search, Shopping, YouTube, Gmail and Display.
What data does Customer Match accept?
Email addresses, phone numbers, first and last names with country and postal code, and mobile device IDs, each normalized and hashed the way Google specifies. Several identifiers per row raise the chance of a match.
Why are B2B match rates lower on Customer Match?
Fewer people sign in to Google with their business email, so business addresses match less often than personal ones. Adding phone numbers and name-plus-location fields, and keeping the list verified and normalized, recovers part of the gap.
Is Customer Match available to every advertiser?
No. It is gated by account policy standing and compliance history, uploaded data must be first-party with appropriate consent, and audiences need a minimum active-member count before serving on some surfaces. Check Google’s current documentation for the exact requirements.
Sources
Every playbook on this blog ships as a runnable mission.
Open a workspace and the playbook library is waiting - describe the outcome and the agents carry it end to end, on your plan's monthly credits.