Customer Match and the Data Manager API: what to do now

Google's developer docs put it plainly: "Starting April 1, 2026, OfflineUserDataJobService and UserDataService requests for Customer Match will fail." That applies to projects with no Customer Match activity from October 1, 2025 to March 31, 2026 (Google Ads API docs). Google's fix: "Use the Data Manager API instead."

Here's Customer Match and the Data Manager API as of November 2026, for a small B2B team.

What changed in the Customer Match Data Manager API move?#

Failed calls return CUSTOMER_NOT_ALLOWLISTED_FOR_THIS_FEATURE. Google Ads Help calls Data Manager "the recommended path forward" (Google Ads Help).

What's different in the new API, per Google's upgrade guide (Data Manager API):

What is Customer Match, and who can use it?#

Customer Match is a Google Ads feature. It "lets you use your online and offline data to reach and re-engage with your customers" (Google Ads Help). You upload emails, phone numbers or addresses. Google matches them to signed-in users on Search, Shopping, YouTube, Gmail and Display.

The Customer Match policy sets two levels, as of November 2026:

Under $50,000 lifetime spend, you can't target a list. You can still exclude it or observe it.

The policy also limits the data. Upload only data "you collected in the first-party context," such as your site, app or store. Your privacy policy must disclose sharing customer data with third parties. No data on users under 13. A bought lead list fails the first rule.

How to hash and format the list#

Get the format wrong and the upload still succeeds, but rows "can't be matched with a customer" (Google Ads API docs). Google's normalization rules:

1. Trim leading and trailing spaces.
2. Lowercase the text.
3. Phones in E.164, such as +14155550100.
4. Gmail addresses: remove dots and anything after a plus sign in the part before the @.
5. Hash email, phone, first name and last name with SHA-256.
6. Don't hash country, zip, IP addresses or timestamps.

In the Data Manager API you also set an encoding field, hex or Base64 (Data Manager API).

Manual uploads are easier. Upload a plain-text CSV and "Google Ads will hash it for you" with SHA-256 (Google Ads Help). Headers must use Google's exact names: Email, Phone, First Name, Last Name, Country, Zip.

What consent do you need?#

Google's EU user consent policy covers users in the EEA, the UK and Switzerland. You need valid consent before using their data for ad personalization, and records of it (Google).

For Customer Match, two consent signals matter: ad_user_data and ad_personalization. Since March 2024, both must be granted for a list to be used in the EEA (Google Ads Help). A missing signal counts as not consented, and that data "will not be processed."

In a manual upload, tick the compliance box and upload only consented data. In the Data Manager API, set adUserData, adPersonalization and termsOfService in the request (Data Manager API). Partners must pass the signals for you. IP matching isn't supported for the EEA, UK or Switzerland. This isn't legal advice.

Which Google Customer Match upload method fits you?#

As of November 2026, Google Ads Help names four routes.

Data Manager's supported sources table includes HubSpot, Shopify, Klaviyo, Google Sheets and BigQuery for Customer Match. My take: start with one file in the UI. Move to a connection once the list earns its keep. Write API code only if no connector fits.

How big should a list be?#

Google gives three numbers:

• 100 rows is the floor for each uploaded file. Google Ads Help also recommends lists of "at least 100 users to prevent ads from not serving" (Google Ads Help).
• 5,000 members is what the API docs suggest "to increase the chance of having enough matched, active users for targeting" (Google Ads API docs).
• 540 days is the longest membership. Lists with too few members added or updated inside that window can stop being eligible (Google Ads Help).

Serving also needs enough active users; Google doesn't publish that floor.

Match rate is the share of your uploaded list that Google can connect to its users (Google Ads Help). Google gives no target range. It says unmatched rows are common, and that match rate "isn't an indicator of list performance." Processing "may take up to 48 hours."

To lift it, Google suggests more match keys, such as email plus phone. B2B lists often hold work emails only, so expect lower rates.

Customer Match for B2B: ABM and exclusions#

Exclude current customers first#

Every eligible account can run this play, even below the $50,000 line. Upload paying customers and exclude them from prospecting. Stop paying for clicks from people who already pay you.

Pull the list from Stripe or your billing tool, not your CRM. Refresh it weekly. Our guide to tracking Stripe conversions in Google Ads shows how to link those payments back to campaigns.

Account-based lists#

Upload contacts at target accounts and use the list in Observation. You'll see whether they search your terms and what they cost. With targeting access, raise bids on them. Pick the accounts with ABM for startups: one account play your team can run in a week. If only a handful match, ads to that list may not serve.

Pair audiences with revenue#

An audience tells Google who to show ads to. Conversion uploads tell it what they paid. Our guide to offline conversion import in Google Ads with real Stripe revenue covers the second half. To judge leads on pipeline, see B2B lead generation strategy: measure pipeline, not form fills.

What breaks for scripts and tools on the Google Ads API#

If you or a vendor built an uploader, check these:

1. Silent failures. A nightly sync that fails quietly leaves your exclusion list stale. New customers keep seeing prospecting ads.
2. Dormant code. A script untouched since last year is the riskiest.
3. Partial-failure logic. The new API fails fast, so one bad row can stop the request. Validate first with validateOnly, which checks "for fast-fail errors without ingesting any data" (Data Manager API).
4. Batch size. Google's sample code sets 10,000 members per request (Data Manager API). Split larger lists.
5. Credentials. The new API needs no developer token; it uses Google Cloud project quotas. Google marks the legacy token "deprecated September 9, 2026" (Google Ads API docs). See Google Ads API without a developer token.

Where Kratly fits#

Kratly's ad audiences, built from a lead list or an account play and synced to the ad platforms, are built but switched off. Google uploads will move to the Data Manager API before we turn them on. Live today: Kratly reads your Google Ads data next to Stripe or Shopify revenue and website tracking, and shows revenue by campaign, page and email, so you can see which campaigns bring paying customers.