Customer Match

Customer Match works by uploading a list of contact data - emails, phone numbers, or mailing addresses, hashed with SHA256 before upload, or synced through an approved CRM or API integration - which Google then hashes on its own side and matches against its signed-in user records. Only the portion of the list that matches an active, signed-in Google account becomes a usable audience; the unmatched remainder isn't a targeting concept at all, it simply doesn't exist as reachable inventory.

Once matched, the list can be used across Search, Shopping, YouTube, Display, and Demand Gen in three distinct ways: as an inclusion, targeting only people on the list (reactivating past buyers, for instance); as an exclusion, suppressing existing customers from acquisition campaigns so you stop paying to reacquire people who would have converted anyway; or as a seed audience feeding Optimized Targeting, giving Google's system a concrete example of who actually converts. That third use matters more than it used to, because Similar Audiences - the original dedicated lookalike feature built on remarketing and Customer Match lists - was fully sunset, and Google has kept consolidating audience expansion into Optimized Targeting rather than a separate standalone lookalike audience type; check current eligibility in the account before assuming a dedicated lookalike-style option is still offered.

Customer Match is strongest for exclusion (protecting margin on branded and acquisition campaigns), win-back efforts targeting lapsed customers, and B2B accounts with enough CRM contact volume to build a meaningful list. It's weak for brand-new businesses with thin customer data, since Google requires a matched list to clear a minimum size threshold before it will activate as a targetable audience at all - a list too small simply won't serve, regardless of how cleanly it was hashed and uploaded.

The most common mistake is assuming the full uploaded list becomes reachable. Match rates run well under 100 percent, because matching depends on the contact having a signed-in Google account using the same email, phone, or address you uploaded, and a list that looks like solid volume on paper can serve to only a fraction of that once matched. A second mistake is uploading sensitive categories - health conditions, financial hardship, and similar data - which violates Customer Match's certification requirements and gets the list rejected or the account flagged for review. A third is treating the upload as a one-time task: contact information ages as people change emails and phone numbers, and lists that are never refreshed shrink and degrade in match quality over time, not just in size.

In reporting, Audience Manager inside Google Ads shows matched audience size directly next to the uploaded file size - if that number looks small relative to the file, any campaign built on top of it will underperform expectations for reasons that have nothing to do with bid strategy or creative quality, and it's worth checking there before troubleshooting anything else. When comparing Customer Match against behavior-based audiences like Remarketing Lists or In-Market Audiences in reporting, remember it's the only one built from data you already own rather than Google's inferred behavior - it typically converts differently than behavioral audiences even at similar reach, because it's targeting known relationships rather than modeled intent.